Using Subordinated Debt as
an Instrument of Market Discipline
Subordinated Debt Discipline
Regulators may use subordinated notes and debentures (SND) in various ways to discourage banks from excessive risk-taking. If such an SND plan is to be effective, then it must impose costs on at least some risky banks at some time. The intent in imposing such costs is to induce banks to change their behavior in ways that reduce their risk of failure or the losses they impose on the Federal Deposit Insurance Corporation (FDIC) if they should fail or both. However, banks may also try to avoid the costs of the SND policy by taking steps that are within the legal framework of the policy but work to defeat the regulatory goals behind the policy. In extreme cases, some banks evade costs by taking actions that are outside the legal framework of the policy.
This appendix addresses the questions of how banks might avoid the regulatory goals of an SND policy and how supervisory and regulatory procedures might deal with this avoidance. The focus is on bank actions that are legal within the framework of an SND plan. The appendix also considers bank actions that induce other market participants to take actions that frustrate the public policy goals of an SND plan. Limited consideration will also be given to actions that are outside the legal framework of the policy.
The objective of regulating individual banks for safety and soundness is generally taken as some combination of the goals of reducing expected losses to the FDIC if the bank should fail and reducing the risk that a bank will fail. The goal of reducing expected losses to the FDIC is desirable in itself for reducing the probability that taxpayers will be expected to cover losses at banks. The goal of reducing losses to the FDIC is also desirable in that the pricing of the current deposit insurance appears to be insufficiently sensitive to the riskiness of various banks. Thus, the existing deposit insurance system may encourage some banks to take additional risk, which increases their risk of failure. Reducing or eliminating expected losses to the FDIC may thereby reduce the risk of bank failure to that which would be observed in the absence of distortionary deposit insurance pricing. If the objective is to reduce the risk of failure to something below that which would be observed without the safety net, then merely eliminating the subsidy to risk-taking is not sufficient.
One way in which an SND requirement may help to achieve the goals of deposit insurance is through the discipline exerted directly by the subordinated debt holders in response to changes in the riskiness of a bank. SND exercise discipline by raising the bank's cost of funds, thereby offsetting some or all of the gains that may flow to equity holders from increased risk exposure. Thus, the extent to which subordinated debt may exercise direct discipline depends on the extent to which it raises a bank's cost of funds. Because SND issues may raise a bank's cost of funds when the debt is repriced, a requirement that banks frequently reprice SND is essential for obtaining this discipline. Furthermore, the effect of SND on a bank's cost of funds depends both on the amount of subordinated debt as a proportion of the risks being borne by all creditors of the bank (including the deposit insurer) and on the extent to which the rate on outstanding SND reflects the riskiness of the bank.1 SND may also facilitate greater direct market discipline to the extent that they reduce the cost of complying with the capital requirements and, thus, permit an increase in required total capital levels.
A second way in which an SND requirement might help achieve the goals is through indirect market discipline exerted by private parties that do not hold SND obligations but that monitor SND rates to determine the risk exposure of a bank. An argument could be made that many banking organizations already issue SND and that market participants may observe the rate paid on these issues; thus, a new regulation encouraging SND issuance would be unlikely to add to indirect discipline. However, an SND policy might stimulate additional indirect discipline in several ways. First, more banks could become subject to this indirect discipline to the extent that the policy induced more banks to issue SND. Second, to the extent that the policy reduced the cost of obtaining SND prices, it might encourage more private-sector participants to use SND prices. Timely SND prices are currently available from investment banks only at a cost that may discourage some potential users from obtaining them. Third, the policy might facilitate comparisons across banks to the extent that it resulted in a further standardization of SND contracts or caused banks to issue SND in more concentrated time intervals. Fourth, the policy might encourage private parties to place greater weight on SND prices by setting regulatory benchmarks for these prices. Private-sector participants are at risk in dealing with a financially troubled bank only if the regulators close the bank or impose other restrictions on the bank during the time of that dealing. Thus, if market participants know that the regulators are using a particular risk measure, then they have an incentive to use the same measure. A good example is the market's emphasis on risk-based capital ratios. The risk-based capital measures are not necessarily good measures of the riskiness of any individual bank, but they are good measures of the probability that the regulators will sanction a bank. As a consequence, banks face significant market pressure not only to remain in compliance with the risk-based capital regulations but also to comfortably exceed the regulatory standard so that the risk of future regulatory intervention is reduced.
NOTE. Larry D. Wall, Research Officer, Research Division, Federal Reserve Bank of Atlanta, Georgia, prepared this appendix. The views expressed are those of the author and not necessarily those of the Federal Reserve Bank of Atlanta or the Federal Reserve System. The author thanks Dan Covitz, Robert Eisenbeis, Douglas Evanoff, and Myron Kwast for helpful comments.
Another way in which an SND requirement might help achieve the goals of deposit insurance is through indirect regulatory discipline exerted by regulators incorporating SND rates into their evaluation of the risk exposure of a bank. The methods for such incorporation of information range from informal use, in which SND rates are primarily an additional source of information, to formal use of the rates as a trigger for some supervisory action. Possible regulatory responses to high SND rates include increased frequency of examination, triggers for prompt corrective action, requiring banks paying high rates to shrink, and requiring banks that cannot issue SND to be closed.
Thus, an SND requirement may induce greater discipline either by influencing a bank's cost of funds or by providing a signal for other market participants or the regulators. Direct discipline exerted by funding costs depends on the amount of debt that is repriced after a bank becomes riskier and the rate the bank must pay on that debt. In contrast, the amount of debt being repriced is, by itself, unimportant to the use of SND prices as a signal for indirect market discipline and regulatory discipline. Indirect market discipline and regulatory use of SND rates depend on the accuracy of the pricing signals obtained from the primary and secondary markets for SND. This analysis suggests that the methods a bank may use to avoid the costs associated with increased discipline depend on the type of discipline the bank is seeking to avoid.
Direct market discipline arises to the extent that a bank's cost of funds increases in proportion to increases in its risk exposure. This fact suggests that a bank may be able to avoid direct market discipline by reducing the debt that is subject to repricing if the bank's risk exposure increases. Reducing the debt subject to repricing can be accomplished by minimizing the total amount of SND that the bank must issue relative to its risk exposure or by minimizing the fraction of the total outstanding SND that must be rolled over at any given time.
Banks may reduce the level of SND relative to their risk exposure by increasing their exposure to risks that are underweighted by the SND requirements and avoiding exposure to risks that are overweighted. The methods for avoiding SND requirements are the same as those that could be used to avoid similarly structured requirements on equity capital.2
StrucEagleTraders.comg an SND policy that avoids the problem of banks' exploiting inaccuracies in the risk measure used to determine the required quantity of SND would not be easy. The problems involved in setting the amount of SND that banks must issue are similar to those associated with fixing the current risk-based capital system. The only difference is that the aftertax cost of SND may be lower than the after-tax cost of equity to a bank; hence a bank's incentive to engage in regulatory arbitrage may be less under a pure SND policy. That is, if SND requirements substituted for risk-based equity requirements, then banks would have somewhat less of an incentive to arbitrage the SND requirements.
A bank may reduce the amount of SND that is subject to repricing by maintaining a sufficient stock of SND outstanding so that it could remain in compliance with the minimum SND standard for a year or more without issuing more SND. In this case, even if rollover provisions required the issuance of some new subordinated debt, the bank could issue the minimum required to satisfy the rollover requirement. To illustrate: A large ($200 billion) bank is required to issue at least $100 million in new debt once a year, but it does not need to issue any more debt to satisfy the regulatory requirements. Even if the bank has to issue the SND at junk bond prices, the effect on the bank's overall cost of funds will be rather small.
The regulators may minimize this form of avoidance by requiring the banks to reprice a substantial amount of SND on a regular basis. One way of doing so, for example, is to require that the bank issue a given fraction of its total SND needs every year. Another way is to shorten the maximum maturity of the SND issues. If a bank may issue ten-year debt and the debt is not subject to the existing discounting, then the bank must maintain SND equal to only 110 percent of its capital requirement to eliminate its need to issue new SND to comply with the yearly minimum SND levels. However, if a bank cannot issue SND with a maturity of more than three years, then it would need to maintain SND levels equal to at least 133 percent of the minimum requirements to minimize its required issuance in any given year. A third way that increases market discipline is to require that all of the bank's SND be repriced on a regular basis. This repricing could be accomplished in various ways, including (1) requiring the debt to be rolled over every period (such as once a year), (2) requiring that the rate on outstanding debt be periodically changed based on observed primary or secondary market prices, or (3) requiring that SND holders be given a put option on the debt so that the bank would have an incentive to reprice the debt regularly in line with the bank's risk exposure.
A bank may avoid all three types of SND-induced discipline to the extent that the actual total price paid by a bank in the primary market does not fully reflect the bank's risk exposure. A bank may also avoid indirect market discipline and regulatory discipline to the extent that it can reduce the observed rate paid on the debt even if it must compensate investors in other ways for the bank's risk exposure.3 However, reducing the observed rate by compensating investors in other ways does not reduce direct market discipline and may even increase the total cost of the debt to the extent that other forms of compensation are less valued by investors.
The key to reducing the actual price paid on a new SND issue is to mislead investors about the actual financial condition of the issuing bank. Banks may issue statements that fail to disclose material exposures or that provide inaccurate or misleading information about its exposure. The buyers of SND generally understand banks' incentives to mislead and would charge a premium for bearing this risk. Moreover, financially strong banks would have an incentive to become more transparent to reduce the risk premium on their SND. Nevertheless, banks still try to mislead investors about their risk exposure and may succeed for a time.4
Various supervisory and regulatory mechanisms already exist to encourage banks to fairly disclose their financial condition because banks currently have some incentive to mislead investors. The principal way in which an SND policy could further reduce this type of avoidance is by increasing bank regulators' efforts to promote transparency. At times, the regulatory agencies have been, at best, ambivalent about the merits of promoting enhanced transparency. The adoption of a plan in which SND play an important role in disciplining bank risk-taking may encourage supervisory authorities to more aggressively promote greater transparency.
Banks may also reduce the observed rate by providing other forms of compensation to investors in the debt. One way of providing such compensation is to structure the debt so that it is more attractive to investors, such as issuing shorter-term debt or embedding valuable options in the debt. However, outside observers, including other market participants and the regulators, would recognize such compensation and could seek to add back the estimated effects of the compensation to obtain the actual cost of the debt. Moreover, the SND policy could be structured in a way that substantially reduces banks' ability to embed compensation in the debt contract.
The other way of reducing the observed cost is to provide the compensation outside the debt contract. Attempts to provide such compensation through an explicit promise by the bank to compensate investors outside the contract would be risky and might constitute fraud. However, such compensation may nevertheless be attempted in various subtle ways. The bank could pay above-market rates on deposits by SND holders, offering below-market prices on transactions services or loans. If the SND issue were being underwritten by outside investment bankers, then the bank could tie its purchases of future investment banking services to the investment banks' willingness to underwrite the issue at a below-market rate. If an affiliated investment banker were underwriting the SND issue, then the affiliate might support the price in the secondary market and provide buyers of the issue with discounted services and a favored position in attractive investments (such as Internet initial public offerings) underwritten by the investment bank.
Supervisory and regulatory steps can be taken to substantially reduce banks' ability to provide compensation outside the SND contract, but completely eliminating such compensation may be impossible. The banking agencies could collect information on the relationships that buyers and current holders of SND have with the issuing bank and its corporate affiliates. These relationships could then be reviewed for evidence of the bank's providing compensation. However, the bank and the investor may not have an explicit agreement, and the compensation may be provided some time after the SND issue is sold. For example, if the regulatory agencies set a trigger price (or rate) on SND issues that will automatically produce some substantial regulatory sanction, a major investor in bank securities is likely to be aware of this trigger point and the benefit the bank gains from avoiding the sanction. Such an investor may be willing to buy part of the issue at a slightly above-market price (below-market rate) to help the bank avoid the regulatory sanction. The investor may do so based not on any explicit agreement but rather on an implicit understanding that the bank owes the investor a favor.5 An additional step that the regulatory agencies can take to reduce the potential for such implicit compensation is to monitor the placement of issues relative to the set of large SND holders. The placement of almost all of a new issue with a few buyers, especially those that rarely invest in SND, could signal that the buyers anticipate compensation outside the SND contract.
A bank may also try to mislead investors by shifting risk outside the bank with the expectation that, if serious problems arise, then the risk may be shifted back into the bank or otherwise covered by the safety net. An argument could be made that such an attempt to exploit a bank SND policy is misguided because the bank has no direct exposure to losses at its separately incorporated affiliates and the supervisors would not permit the bank to assume risks from its affiliates. This argument may have some merit with respect to affiliates whose operations are independent of the bank. The argument has less merit in those cases in which the bank and its nonbank affiliate are marketing a package of bank and nonbank services. In this case, the failure of the nonbank affiliate may have a significantly adverse effect on the bank even though the bank is not technically liable for the affiliate's losses. However, SND investors should recognize that the bank may have some exposure to its nonbank affiliates and incorporate this risk into the price of an SND issue. Thus, increased reliance on bank-issued SND may help in addressing one of the more difficult problems faced by regulators-how to evaluate the implications of nonbank affiliates for the safety and soundness of banks.
Secondary market rates and prices may be important for indirect market discipline and regulatory discipline based on SND rates. Thus, banks may avoid these types of discipline by reducing the observed rates paid on SND to a level below that which reflects the riskiness of the bank. Both mechanisms for reducing the observed rate on primary issues may also be used to reduce the observed rate in the secondary market. However, reducing the observed rate by other forms of compensation may be both easier and more expensive to do. It may be easier because a smaller fraction of the SND issues is likely to be sold in the secondary bond market. A significant fraction of this debt is likely to be purchased by investors after a buy-and-hold strategy. Moreover, because some investors may be unable to readily observe market prices, they may not realize that the secondary market price is being artificially inflated and, thus, may keep their SND claims because they believe that they would receive only fair-market value for their holdings. However, any additional compensation the bank does offer to purchasers in the secondary market will add to the cost of the issue. That is, when the bank compensates primarymarket investors for receiving a below-market rate, the bank is merely changing the form of the compensation, substituting a lower rate on SND for morefavorable terms on some other product. After the debt is issued, however, the bank is under no obligation to increase the total compensation provided to the holders of the debt. Yet when a bank provides implicit compensation (such as loans at a belowmarket rate), it is effectively increasing its payments to at least some debtholders.
The observed rate in the secondary market could also be reduced if the dealers in the bank's SND issues reported inflated prices to the regulators. The SND market is a dealer market with no central collecting of transactions prices and a notable reluctance on the part of dealers to provide actual transactions prices. The only effective check the regulators have on the estimates provided by dealers may be the actual transactions prices of primary issues. Dealers may expect to receive some compensation for providing biased estimates, perhaps in the form of implicit understandings about future dealings with the investment bank. Given the low cost of providing misleading information to the regulator, the required compensation for doing so may not need to be very large. Indeed, investment banks already seek to win business from bank and nonbank firms by having their stock analysts produce favorable reports.
Supervisors could seek to minimize investment bankers' incentive to report inflated prices by obtaining actual transactions prices from several dealers. Supervisors might also compare reported secondary market prices with new issue prices to obtain information on the quality of the estimated prices obtained from investment bankers.
Finally, observed rates in the secondary market could be reduced by placing a large fraction of the more actively traded issue or issues with a single investor.6 Often the supervisory interest in protecting the safety net is consistent with the SND holders' interest in receiving the promised payments on the SND issue. However, if a bank is sufficiently distressed, then the bank's and the SND holders' interests may diverge. A small investor with adverse information may be able to liquidate most or all of its position before triggering increased indirect market and regulatory discipline. An investor with a very large position may not be able to sell a significant portion of the holdings before triggering increased discipline.7
Supervisors may minimize this risk by monitoring the fraction of each bank's SND held by the largest investors. A small number of investors holding a large fraction may signal potential problems, particularly if these investors are purchasing SND on the secondary market and the rate paid on the SND is close to the regulators' trigger point. Also, weak banks may benefit from a concentration of SND ownership, but stronger banks may be threatened by such a concentration. If SND are concentrated in a few holders, these holders may be able to blackmail the bank by threatening to dump their holdings on the market and significantly reduce the market price. Thus, healthy banks may seek to ensure that their SND are widely distributed when they are sold in the primary market.
The prices of SND issues may move significantly in a short time. In some cases, a large change in price may reflect a realistic reassessment by the market of a bank's prospects. However, in other cases the magnitude of the movements may appear to be out of proportion to any fundamental news about the bank. Furthermore, the liquidity in the SND market may plunge to levels that call into question the accuracy of any price quotes. For example, market participants have stated that SND prices fell and liquidity dropped substantially in September and October 1998 after the Russian bond default. Price movements that appear unrelated to fundamentals and periods when the SND market has minimal liquidity potentially challenge the use of SND to generate market discipline and to signal the regulators. The implications of large price movements and reduced liquidity depend on whether the shock is influencing (1) only a specific bank or group of banks or (2) the entire bond market (bank and nonbanks).
One shock that may affect a specific bank or banks is false rumors about them that significantly reduce SND prices. Two problems could be associated with false market rumors about specific banks. First, banks themselves could try to exploit the vulnerability of SND prices to rumors by starting false rumors to discredit regulatory use of SND. However, it is not clear that a bank would gain from starting rumors about itself. If the rumors were quickly dispelled, then the harm from the drop in SND prices would be minimal, and regulatory confidence in their long-term value might not be reduced. If the rumors were not quickly dispelled, then they might have a greater effect on regulatory confidence, but they might also have a negative effect on the bank's dealings with some of its customers and suppliers independent of any change in SND prices.8 In practice, banks may find it safer to wait for others to start the rumors.
The other problem arising from rumors is that the bank could be forced to issue new subordinated debt or that the regulators would respond to an SND rate trigger before the rumors were dispelled. However, the effect of such rumors should be short-lived. Banks could offset the impact of rumors by providing additional information to the market. In one of our interviews, we were told that the interviewees' firm had been the subject of false rumors and that the firm had countered the rumors by providing additional information to reassure market participants. Banks may prefer to keep some of this information confidential. However, the risk of being forced to reveal confidential information is greatest for financially weak banks. Thus, the banks that are most at risk of having to make costly disclosures are those that, from a safety and soundness perspective, the regulators would most want to bear this risk.
Another source of noise in SND risk measures is that the market may receive valid information, but the implications of that information for specific banks may be unclear. For example, if a very large borrower experiences financial distress, then the prices of the SND of all potentially significant creditors may decline. In this case, SND traders are likely to be worried that the other side in their transaction knows more than they do. Thus, traders will tend to post very large bid-ask spreads to protect them from losses arising from trades with informed investors.9 Given large spreads, investors that do not have superior information may be unwilling to trade. Further, if an investor is willing to transact at the posted prices, the dealer may take that as a signal that the investor has superior information and refuse to trade even at the posted price.
That bank regulators may not rely on SND prices as a risk signal during periods of adverse rumors or illiquidity raises the possibility that a bank could try to exploit this situation by taking on additional risk when its SND are illiquid. The debt may not remain illiquid for long, so this risk appears remote. A potentially more significant problem is that a bank may be required to issue SND during a period of illiquidity or that a ''false'' price signal from the SND market could trigger a regulatory response.
The possibility that rumors or incomplete pieces of information will significantly influence a bank's SND prices has several implications for the structure of an SND policy. Any policy that requires banks to issue SND during a short interval risks imposing costs on the bank that serve to provide no safety and soundness benefits. Banks may be able to counter false rumors or supply missing information about actual exposures to new sources of loss if given a few days. Thus, banks should generally be given some period of time within which to issue SND.
However, giving banks a window within which to issue SND raises two potential problems. First, banks may seek to time issues during ''good'' periods so that the prices of primary issues are not a random sample of the bank's financial condition. The timing of bank SND issuance could be a problem to the extent that the regulators rely heavily on prices from infrequent primary market issues. This seems less of a problem to the extent that secondary market prices are also used, as they would for larger, longer-term debt, or that banks are required to make frequent primary market issues, as may be the case for smaller, shorter-term debt issues. Second, banks may try to exploit regulatory concern about the impact of rumors and incomplete information by deferring the issuance of new debt until the end of their window and then hoping that a market disturbance will allow them to further defer such issuance. Regulators may counter this incentive to defer issuance by following a policy of rarely granting permission to defer issuance because of bank-specific market disturbances. If banks know that regulatory permission to defer their issuance of SND will rarely be granted, then they will generally try to complete their SND issue during the early to middle parts of the window.
If the regulators are using SND prices as a signal, then automatic triggers based on a single day's price may not be desirable. Instead, a large jump in the SND rates paid by a bank might trigger the supervisors' second look at the bank to determine whether the price contains new information. However, if the intent of an automatic trigger is to prevent lengthy forbearance, then the regulators should give the bank some time (a few days to two weeks) to tell its story to the market and correct any mispricing.
The implications of a large move in SND prices are different if changes in the bank SND market mirror changes in the overall bond market. Sharp increases in the spreads of all corporate debt over comparable Treasury securities have been called flights to quality and are sometimes explained as an increase in risk aversion. This overall increase in risk premiums may be due to information about the economy, but it may also reflect an increase in market concern about the probability of an event that has not yet occurred.10
Neither banks nor bank regulators may be able to counter sharp increases in overall corporate bond spreads over Treasuries. Nonfinancial corporations will generally defer long-term bond issuance during such a period, preferring to wait for spreads to narrow. Requiring banks to issue long-term SND during this period may impose substantial costs on them. One alternative is to grant forbearance to banks that defer issues for a fixed time when overall corporate bond spreads over Treasuries exceed a given threshold. Such a policy, however, may take away the SND signal just when such a signal may be especially valuable. To prevent banks from exploiting the temporary absence of an SND signal, another alternative is to require banks to issue subordinated obligations but allow them to shorten the maturity to one year or less. Permitting banks to issue shorter-term obligations would have two beneficial effects: (1) Shorter-term obligations are generally likely to have smaller risk premiums, and this difference may be more pronounced during periods of flight to quality, and (2) shorter-term obligations can then be refunded at lower rates when market concerns dissipate. However, allowing banks to issue shorter-term obligations implies that the signal from the SND market is likely to be more focused on the short-term prospects of the bank than it would be with longer-term issues.
The potential for a general increase in corporate bond spreads argues against setting SND rate triggers for regulatory action at fixed spreads over Treasury securities. Any trigger should be based on a measure that adjusts for overall risk premiums. Examples of such measures include bank SND rates relative to corporate bonds of a given rating (such as Baa) and converting SND spreads by the swap curve to a premium over libor.
The preceding analysis suggests several ways in which banks may seek to avoid SND-induced discipline (see table C.1 for a summary of these methods). However, virtually any plan that seeks to create a substitute for adverse effects of the safety net on market discipline is going to have some weaknesses. The relevant question is whether any SND plan could be effective given banks' alternative methods of avoiding this type of discipline.
Most banks may engage in only minimal efforts to undercut SND discipline most of the time. Unfortunately, this point is of limited comfort because banks that are experiencing financial difficulty are those that we are most concerned about and that are most likely to seek to avoid SND discipline.
The potential problems with using SND to obtain direct market discipline are substantial, particularly if the plan must operate within the rules of the 1988 Basel agreement. Many observers are exploring alternative methods of inducing greater discipline precisely because of weaknesses in the existing risk-based capital measure. Moreover, our current five-year discounting period results in bank SND issues typically having ten-year maturities. As a consequence, at worst a bank might need to roll over only 10 percent of its SND requirements, which may be limited to 2 percent of risk-weighted exposure under the Basel agreement. The bottom line is that the bank may need to reprice only 0.2 percent of its risk-weighted exposure in any given year, and the risk-weighted exposure measure almost surely understates the actual risk of a financially weak bank.
Although banks may have several ways of reducing the observed rate paid on their SND issues, this problem may not be insurmountable. Banks might be able to reduce their observed SND rates a few basis points by offering compensation outside the SND contract. However, banks that sought to reduce their observed SND rates enough to place them in a higher rating category (for example, to move the rate paid by Ba2 firms to the rate paid by Baa3 firms) would have difficulty doing so in the United States.11 Investors, the rating agencies, and the supervisors
would likely suspect avoidance activity by the bank, and the bank would risk incurring additional supervisory sanctions for such avoidance. Moreover, to the extent that the regulators may anticipate that banks will be able to reduce observed SND rates, they may also reduce their trigger point for taking supervisory or regulatory action. Finally, the incentive to mislead regulators may be reduced by imposing discipline in a series of smaller regulatory actions rather than in a few draconian measures. Such a continuous response reduces the gains from moving observed SND rates a few basis points.
Probably the biggest concern with using SND rates for indirect market discipline and supervisory discipline is that banks will temporarily mislead investors about their true risk exposure. This problem is not unique to an SND policy, however; it may occur in any attempt to use market forces to discipline bank risk-taking. Probably the best supervisory response to inadequate transparency-besides continuing efforts to enhance it-is to maintain bank supervisors' ability to act independently of the SND signals.