Using Subordinated Debt as
an Instrument of Market Discipline
What Characteristics Should the Required SND Have?
Advocates of an SND policy have proposed many specific characteristics that an SND instrument should have to meet particular policy objectives. This subsection examines what the study group judges to be the most important of these instrument characteristics.
If an SND policy is to increase market discipline, it is virtually essential that the SND instrument be tradable, or issued quite frequently, in a competitive market to independent third parties. This is a requirement of virtually all SND policy proposals. 53 Only in this way could the primary policy objectives be achieved. Put differently, SND issued to insiders of the bank or bank holding company would clearly subvert the incentive and price-signaling objectives, although they could provide extra protection for the FDIC. 54 The fact that a large and liquid market already exists for the SND of the major bank holding companies and for some of the major banks demonstrates that requiring the largest banks to issue tradable securities is operationally feasible.
According to study group interviews with market participants, about eight major independent investment banks are the most important underwriters and dealers of bank and BHC SND. In addition, some Section 20 subsidiaries of large bank holding companies were said to play a role. Indeed, one interviewee suggested that Section 20 subsidiaries sometimes purchase their holding company's SND to support its price. The study group had no way of substantiating this claim, but in principle there may sometimes be an incentive for Section 20 subsidiaries to engage in such behavior, and an SND requirement, especially if it included a rate cap, would increase that incentive. 55
The incentive for a Section 20 holding company subsidiary to support the price, or otherwise subsidize the SND, of its bank affiliate or holding company parent is a potentially serious impediment to achieving the objectives of an SND policy and is an argument for prohibiting such activity. Indeed, this concern was discussed in the context of a more general consideration of potential conflicts of interest when the Federal Reserve Board was considering Section 20 subsidiaries in the 1980s. In light of these concerns, regulatory firewalls were established that prohibited such activity.
In August 1997, the Board rescinded these firewalls on the grounds that the National Association of Securities Dealers (NASD) Rule 2720 imposes essentially the same restrictions. Rule 2720, to which Section 20 subsidiaries are now subject, provides that if a member of the NASD proposes to underwrite, participate as a member of the underwriting syndicate or selling group, or otherwise assist in the distribution of a public offering of its own or an affiliate's securities, then (1) the securities must be rated by a qualified, independent rating agency, (2) the price or yield of the issue must be set by a qualified independent underwriter, who shall also participate in preparing the registration statement and prospectus, offering circular, or similar document, exercising due diligence, or (3) in the case of equity securities only, there must be an independent market in the security.
Besides these rules and regulations, there are economic incentives that deter a banking organization from artificially maintaining the price of its own securities. Rational market participants are aware of the incentive for Section 20 subsidiaries to manipulate the price of their organization's securities and should discount the price at which they are willing to buy from such a subsidiary its ''own bank'' securities, including SND. Thus, to preserve its own business reputation and to allow its SND to trade, a banking organization should be reluctant to support the price of its SND.
Despite these regulatory and economic safeguards, it is unclear whether the incentive to support would dominate, and the answer may well vary with market conditions. For example, during normal times the benefits from price-supporting behavior may be considerably smaller than the costs of doing so. But in times of financial stress, the temptation could be great to support the price of an affiliate's or a parent's SND, especially if the SND included an interest rate cap. Most important to the consideration of achieving an effective SND policy, times of financial stress are precisely when the market-disciplining effects of SND would be most powerful and most desirable.
Allowing banks and bank holding companies to hold in their investment and trading portfolios their own affiliates' and subsidiaries' SND would be inconsistent with maintaining and encouraging a competitive market for bank and BHC SND. Thus, a strong case exists for prohibiting or, at a minimum, limiting severely such holdings. For example, it might be acceptable to limit holdings of ''own firm'' SND only to amounts that do not qualify for the minimum required by an SND policy. Put differently, it would be desirable to limit SND that would qualify as acceptable under an SND policy to those that were not held in the portfolio or trading accounts of any affiliated entity.
Study group interviews indicated that virtually all recent bank and BHC SND have an initial maturity of ten years. Some interviewees argued that the ten-year standard maturity is driven in large part by the requirement in the Basel Accord and in the banking agencies' capital rules that to qualify as risk-based capital the SND must be amortized on a straight-line basis over the five years preceding its maturity. That is, 20 percent of an SND issue is disqualified from inclusion in tier 2 capital for each of the last five years before maturity.
The standard ten-year initial maturity is an important element of the homogeneity of the current market for bank and BHC SND. This homogeneity eases the interpretation and comparison of secondary market yields, and their easing, in turn, facilitates both the direct and the indirect market discipline roles of SND. In addition, current market participants are obviously familiar and comfortable with the ten-year format. Thus, continuing the convention of the ten-year initial maturity would presumably help to minimize any additional costs incurred by market participants, including regulators, if an SND policy were implemented. All of these arguments suggest that adopting an SND policy that preserves the standard ten-year initial maturity has considerable appeal.
Study group interviewees also frequently maintained that shorter-maturity bank and BHC SND would be issued and demanded by investors if the five-year amortization schedule were relaxed. Market participants argued that the three-year and five-year maturity bond markets were particularly deep and would be attractive to banking organizations. From the point of view of achieving the objectives of an SND policy, shorter maturities would still impose direct and indirect market discipline and would augment direct discipline if issuance became more frequent. More-frequent issuance would also be likely to improve the quality of the price signal, and therefore the indirect market discipline provided by SND. The only real concern would be if the SND became so short-term that it was ''runable,'' but this possibility seems remote. Encouraging varying maturities would, however, obviously complicate matters and cause some disruption to the existing market-a market that has many advantages from the point of view of achieving the objectives of an SND policy. Indeed, preserving the existing ten-year initial maturity while implementing a policy that expanded the supply of tradable SND would likely help to improve the liquidity, depth, and overall efficiency of the current market. Allowing for shorter maturities would also require changes in the existing Basel Accord if we intend for SND to go toward satisfying an SND requirement and, at the same time, to count completely toward tier 2 capital. The reason is that the Basel Accord amortizes SND with remaining maturities of five years or less.
According to market participants interviewed by the study group, the standard SND instrument currently issued by banks and bank holding companies has no call options attached. Previous research and data collected by the study group suggest that call options were not an unusual feature of SND in the middle to late 1980s. In recent years, call options have fallen out of favor. Some market participants indicate that call options are less common today because institutional investors have wanted to lock in their yields and have attached very high prices to calls given the recent backdrop of declining interest rates.
In regard to SND policy considerations, the lack of a call option facilitates the interpretation of price changes and interest rate spreads observed in the market. Thus, the absence of options clarifies the primary signal on which direct and indirect market discipline is based. In short, the existing welldeveloped market for BHC (and some bank) SND has evolved to the point at which call options are not usually attached to the SND, and from the point of view of implementing an effective policy, one has no reason to encourage their development.
The arguments for and against put options are less clear. As discussed in section 1, some advocates of an SND policy have proposed attaching a put option to strengthen (1) market discipline by giving the SND holders a strong say (and perhaps even control) over when the issuing bank would be closed and (2) supervisory discipline by encouraging supervisors to promptly resolve troubled banks. These virtues of a put feature, however, are potentially also its vice. The put option would inevitably complicate the interpretation of price signals provided by SND. Perhaps more important, the put option could nullify, or at least complicate greatly, bank supervisors' ability to choose when to close a bank. Putting aside the consideration of whether attempting to remove the closure decision from supervisors' hands is realistic, giving the closure decision to decentralized SND holders could easily be pro-cyclical because many banks tend to have financial difficulty as macroeconomic activity declines. Given the observed positive correlation of risks across many banks, the nearly simultaneous exercise of put options across many banks could exacerbate a situation with potential systemic risk implications much as a bank run would. Although the threat of a bank run provides strong market discipline, introducing such a threat as part of an SND policy seems problematic at best and is inconsistent with the ''nonrunable'' benefit of SND.
According to study group interviews, the typical U.S. bank or BHC SND instrument is a fixed-rate security. 56 Banking organizations usually swap the interest payments on these fixed-rate bonds for floating-rate payments tied to libor in order to better match the interest flows on their assets. As with the ten-year maturity and the noncallability of SND, the fixed-rate characteristic is an important element of the homogeneity of the current market for bank and BHC SND. Thus, for all of the reasons discussed in previous portions of this study, maintaining the fixed-rate nature of SND would help to ensure the success and minimize the costs of an SND policy.
The recent SND policy proposal by Calomiris would set a maximum yield spread over comparable Treasuries (he suggests 50 basis points) or a ''rate cap.'' Under the proposal, SND could not be issued at yields over that spread. The purpose of this provision is to impose a clear penalty on highly risky banks: ''Banks that fail to roll over their debts at or below the mandated yield spreads eventually would have to contract their risk-weighted assets to remain in compliance'' with the minimum subordinated debt requirement. 57 In the limit, a highly risky bank could be forced to close. If truly enforced, a rate cap could be quite effective for preventing a widespread deterioration of the banking system's assets, such as occurred in the 1980s.
On the down side, a rate cap on bank or BHC SND would almost surely be pro-cyclical, perhaps in a highly damaging way. As SND risk spreads widened during an economic downturn, banks would have to take actions to offset the impact on their perceived riskiness. These would include tightening lending standards and asset portfolio components in a pro-cyclical manner and shifting funding away from deposits and toward equity or SND. Although any binding capital (or other supervisory) policy tends to be pro-cyclical, the macroeconomic implications of a 2 percent to 3 percent SND requirement would be expected to be minor, so long as the policy did not include a rate cap (see appendix D). The need to take pro-cyclical steps in the face of an actual or anticipated downturn would be particularly imperative if the rate cap were defined relative to a riskless rate. Also, a rate cap could become binding during illiquid bond markets even for ''safe'' banks, and thus the constraint could exacerbate a liquidity squeeze on the corporate sector, with potential macroeconomic consequences.58
These potential consequences suggest that a lessrestrictive approach would be to define the SND spread relative to some private rate, perhaps the average rate on the SND of a bank's peer group. However, this approach would be aimed more at identifying individual banks that the SND market viewed as highly risky than at limiting risks undertaken by the entire banking system because, if all of a bank's peers were very risky, relative spreads might show little change.
Even if a rate cap were deemed desirable in principle, the problem of determining the optimal level of the cap would remain. If the maximum spread were too wide, it would have little or no effect on bank behavior. But if the maximum spread were too small, banks would be prevented from taking the prudent risks that lie at the core of their economic function. If the maximum spread were based on SND rates among a bank's peer group, too narrow a limit might suppress some healthy diversity among banks. A single rate cap for all institutions would surely introduce significant behavioral distortions; but determining a unique cap for individual institutions seems a practical impossibility. Moreover, even assuming an optimal cap could be determined for use under normal market conditions, such a cap could lead to major banking disruptions during a period such as the U.S. financial markets experienced in the fall of 1998. This argument suggests that the spread limit might best be defined on an average or moving average basis or that a mechanism would be needed to allow its suspension in the presence of extraordinary market shocks. Addressing all of these concerns would greatly complicate an SND proposal.