4 Key Fixes to the Banking Agencies’ Long-Term Debt Proposal

In an effort to further enhance the resilience and resolvability of large banks, the federal banking agencies are proposing to require them to issue a minimum amount of unsecured long-term debt.[1] The proposal can be expected to produce some benefits, but it is over-calibrated and unnecessarily complex, driving up costs in excess of its benefits. Four essential fixes could remedy the problem.[2]

Rationale Behind the Proposal

By requiring banks to issue and maintain a minimum amount of long-term debt on their balance sheets, the proposal would ensure that each bank has an extra layer of loss-absorbing resources on top of the bank’s existing capital requirements. The extra layer of protection is intended to yield three main benefits. First, it should help bolster depositor confidence in the bank during periods of stress and thereby mitigate the risk that uninsured depositors might panic and withdraw their funds en masse. Second, in the event that the bank does fail and is placed into an FDIC receivership, the extra layer of loss protection should increase the FDIC’s options for resolving the bank in an orderly way. Third, as a result, it should reduce losses to the Deposit Insurance Fund (and thus, it follows, should reduce the size of assessments imposed on the banking industry).

It makes sense that the agencies would seek to improve the resolvability and resilience of large banks following the March 2023 failures of three larger banks. But, as proposed, the rule has fundamental flaws that would impose outsize costs on regional banks.

Proposal’s True Costs Far Greater than Agencies’ Estimate

BPI research has shown that the proposal is likely to be more than three times costlier than the agencies estimated in the proposed rule.[3] BPI’s projections are likely more reliable. As detailed previously, BPI considered several factors that were not considered in the agencies’ analysis, including the effects of shortfalls at the subsidiary bank level, the necessity of replenishing the liquidity coverage ratio (LCR) at the holding-company level, the possible rise in risk-weighted assets (RWAs) stemming from the Basel proposal (which would, in turn, raise long-term debt requirements) and the need for banking organizations to maintain management buffers to avoid dipping below minimum requirements due to business-as-usual fluctuations in balance-sheet size or market downturns.[4] Additionally, while the agencies used certificates of deposit data for only six of the 20 entities covered by the proposal,[5] BPI’s estimates relied on more comprehensive data consisting of credit spreads for individual bonds for 17 of the 20 entities.[6]

Thus, for example, the agencies estimate a total funding cost of $1.5 billion, while a more accurate estimate is $4.9 billion.[7] Further, these costs are likely to burden smaller regional banks most significantly because BPI estimates the proposal would increase pre-tax annual funding costs by $3.0 billion on average for banking organizations with total assets between $100 billion and $250 billion (Category IV banks), compared to the $1.8 billion increase for the larger banks subject to the proposal, in part driven by the higher amount Category IV banks would likely have to pay investors for their issued debt.[8]

Unnecessarily high costs would divert bank funds from valuable economic activities, such as providing credit to U.S. small- and medium-sized enterprises. It is therefore critical that the agencies design the rule to achieve its regulatory objective in the most targeted way possible, while providing banks with flexibility to manage their balance sheets efficiently.

It is impossible to know just how costly the proposal will ultimately be, in part because the agencies are simultaneously considering other regulatory changes that are intertwined with the proposed long-term debt requirements. One of the factors contributing to the proposal’s heightened costs is that it overlaps with the agencies’ Basel capital proposal. The long-term debt proposal is calibrated based on RWAs,[9] and the Basel capital proposal may increase RWAs by $2.2 trillion.[10] Until a capital rule is finalized and implemented, neither the agencies nor the public will have a complete picture of the overall loss-absorbency requirements applicable to larger banks or the information necessary to fully assess the long-term debt proposal’s real benefits and costs. It also is unclear whether the agencies considered the potential effect of a nearly 25 percent increase in annual issuance levels of long-term debt on investor demand, pricing or credit spreads.

4 Key Fixes

Four key fixes would help to rein in the proposal’s outsized costs while preserving its intended benefits:

1. Recalibrate to a level more appropriate for regional banks

The proposal would require a covered bank to hold debt equal to at least 6 percent of its risk-weighted assets.[11] But this amount of long-term debt is unnecessary to accomplish the goal of improving the resolvability of a regional banking organization.

This 6 percent requirement is based on a “full capital refill” concept, which is borrowed from the long-term debt requirement in the TLAC rule applicable to GSIBs.[12] The full capital refill concept was created to fully recapitalize, and thereby allow to continue operating, key holding company subsidiaries by downstreaming funds from their parent holding company. The principal focus of this concept was on broker-dealer subsidiaries, which tend to be short-term wholesale-funded and created the greatest too-big-to-fail risks during the Global Financial Crisis.

Regional banking organizations, on the other hand, typically hold the great majority of their assets in their subsidiary banks or retail brokerage subsidiaries that could be sold or wound down in resolution proceedings. They have typically have bank resolution strategies premised on the bank entering FDIC resolution proceedings. In FDIC resolution, the bank would need to be recapitalized at a level that is sufficient to give the FDIC enough time to execute its resolution strategy successfully, minimize losses to the DIF and otherwise provide the FDIC, as receiver, with incremental flexibility in resolving the bank.

A 2 percent calibration would be a more reasonable level of long-term debt to require for regional banks with resolution strategies that are premised on the bank entering resolution proceedings. A 2 percent calibration is based on a more suitable set of assumptions: (i) that the bank would not continue to operate, thereby negating the need for a stress capital buffer of 2.5 percent of RWAs, and (ii) that a distressed bank would be likely to enter receivership before its going-concern capital is fully depleted.[13]

This 2 percent calibration would reduce the cost of the proposal while still providing additional loss-absorbing capacity to improve the resolvability of regional banks and reduce the costs of their failure.

2. Eliminate the overly prescriptive internal issuance requirement

The proposal would require a covered bank to issue qualifying long-term debt internally to its holding company, which would, in turn, be required to issue debt to the market.[14] This highly prescriptive “internal issuance” requirement produces higher costs in significant, and likely unintended, ways.

First, the internal issuance requirement would almost certainly increase the overall amount of long-term debt a bank needs to issue at the holding company. In the most basic example, suppose a bank holding company currently has long-term debt equal to 3 percent of RWAs. The proposal would allow existing debt to meet the new requirement, which would mean the holding company would need to issue another 3 percent of debt. However, if the holding company does not currently downstream those funds in the form of internal long-term debt (e.g., because the holding company uses those funds for other corporate purposes), it might still need to issue the full 6 percent to have the funds available to downstream to the bank.

More particularly, because the rule specifies that the bank-level funding must be in the form of eligible internal debt, that means existing funding mechanisms, like intercompany deposits, would not count. Although a bank could theoretically convert existing internal funding sources into eligible internal long-term debt, BPI research has demonstrated that banks cannot simply convert existing intercompany deposits into internal long-term debt without consequences for liquidity regulation. In particular, intercompany overnight deposits increase the amount of HQLA available at the bank level to meet the holding company’s LCR, and banks may need to issue more debt to maintain the LCR at the holding company.[15] The proposed rule did not consider this complication.

The agencies can restore the calibration to an appropriate level by eliminating the internal issuance requirement and allowing banks more flexibility to manage their internal funding. The agencies can provide greater flexibility by treating other sources of funds as eligible long-term debt, such as intercompany deposits, high-quality liquid assets pledged by the holding company to secure its obligation to recapitalize the bank or other funding mechanisms approved by the Federal Reserve and FDIC in resolution planning. These additional forms of loss-absorbing capacity would continue to provide additional resources at the bank level to enhance resolvability and minimize losses to depositors while reducing the hidden costs arising from a prescriptive internal issuance requirement.

3. Tailor requirements as mandated by law

As noted above, the proposal would extend a 6 percent long-term debt requirement to all banks with assets over $100 billion and would present proportionately higher costs for smaller regional banks. In total, Category IV banks would face a long-term debt shortfall of $104 billion, while larger banks subject to the proposal would face a shortfall of $83 billion.[16]

This result is exactly the opposite of what was intended by Section 165 of the Dodd-Frank Act, which requires the Federal Reserve to differentiate the application of enhanced prudential standards based on a bank holding company’s size, capital structure, risk, complexity, financial activities or other risk-related factors.[17] The Federal Reserve previously recognized a similar rule, the TLAC rule, as an enhanced prudential standard under Section 165.[18] Section 165 also requires the Federal Reserve to determine whether applying such enhanced prudential standards to Category IV banks is necessary to prevent or mitigate risks to U.S. financial stability or to promote safety and soundness.[19] The Federal Reserve has not done so, even though the proposal would have a particularly significant effect on Category IV banks. This proposal presents a perfect case for tailoring because, as noted above, long-term debt serves a different purpose for these banks. Rather than supporting a resolution strategy based on a full recapitalization of material subsidiaries, the stated purposes of long-term debt in the current proposal is to provide the FDIC incremental flexibility in resolving the bank and providing greater loss absorbency for the DIF, objectives that can be served by a lower, tailored calibration.

To address this issue, the Federal Reserve should differentiate long-term debt requirements for large banking organizations in accordance with its statutory responsibility. Tailoring the proposal based on the statutory factors would help to right-size the costs of the proposal while also adhering to the letter and spirit of the law.

4. Eliminate the minimum denomination requirement

The proposal would require that long-term debt be sold to investors in minimum denominations of $400,000.[20] The proposal would also extend this requirement to the TLAC rule applicable to GSIBs. The agencies’ stated objective is to set a minimum denomination high enough that ordinary individuals aren’t likely to purchase the debt, because the agencies do not want distressed banks to be “disincentivize[d]” from using the debt out of reluctance to impose losses on retail investors.[21]

However, a $400,000 minimum denomination is too high even for institutional investors.[22] The industry-standard minimum denomination is $2,000, and more than 90 percent of trading activity occurred in sizes less than $400,000 in an analysis of trading activity over a recent, ordinary four-week period.[23] By some estimates, less than half of key institutional investors are large enough to make such a large allocation to the debt of a single banking organization,[24] given their need for portfolio diversification.[25] So, the proposal would significantly limit demand for long-term debt while increasing its supply. The agencies do not appear to have considered the key impacts of the minimum denomination requirement as part of their economic impact analysis.

Furthermore, direct investment by retail investors in U.S. banks’ long-term debt has historically been very limited. Retail investors tend to invest directly in securities that are listed on a securities exchange, and, like the vast majority of debt securities, substantially all U.S. bank long-term debt securities are not listed. Additionally, unlike in certain non-U.S. jurisdictions where regulators have the authority to write down the full amount of debt and then offer debtholders shares in the institution,[26] any losses imposed in the U.S. would occur under the protection of applicable insolvency law.[27] In any event, the retail investors who purchase long-term debt securities will have received disclosure about the resolution-related risks, and the federal securities laws do not restrict issuing securities to various classes of prospective investors based on the loss-absorbing characteristics of those securities. For these reasons, the agencies should eliminate the minimum denomination requirement, which would raise the price of all banks’ eligible long-term debt securities without a commensurate benefit to retail investors.

Conclusion

To be sure, there are other issues in the proposal that need to be addressed.[28] But if the agencies were to adopt these four essential fixes, it would go a long way toward addressing the proposal’s deficiencies. Enhancing resolvability and resilience are worthwhile objectives. They deserve a reasonably calibrated, operationally feasible, rationally marketable and lawfully tailored method to achieve them.


[1] See Long-Term Debt Requirements for Large Bank Holding Companies, Certain Intermediate Holding Companies of Foreign Banking Organizations, and Large Insured Depository Institutions,88 Fed. Reg. 64,524 (Sep. 19, 2023). The proposed requirement would apply to banks and bank holding companies with assets of more than $100 billion. It would also make certain changes to the existing requirements already in place for GSIBs, which must comply with a separate total-loss absorbing capacity (“TLAC”) rule and an existing stand-alone long-term debt requirement for resolution purposes. See Total Loss-Absorbing Capacity, Long-Term Debt, and Clean Holding Company Requirements for Systemically Important U.S. Bank Holding Companies and Intermediate Holding Companies of Systemically Important Foreign Banking Organizations, 82 Fed. Reg. 8,266 (Jan. 24, 2017). Thus, it would primarily impose requirements to issue new securities on regional banks, though it is worth noting that the internal issuance requirement discussed below would also apply to U.S. bank subsidiaries of foreign banking organizations.

[2] While there are many other changes the agencies should make to improve the rule in addition to those listed below, these fundamental changes to the rule’s calibration, design and application are key to right-sizing the proposal’s effects on the institutions it is intended to strengthen.

[3] See Haelim Anderson, Francisco Covas and Felipe Rosa, The Long-Term Debt Proposal and Bank Profitability, Bank Policy Institute (Dec. 7, 2023) (link).

[4] See id.

[5] See 88 Fed. Reg. at 64,552.

[6] See Anderson et al., supra note 3.

[7] See id. Additionally, the agencies estimate a 3-basis-point decline in net interest margin (NIM); BPI finds a 10-basis-point decline in NIM. The agencies’ assumptions would lead to a 26-basis-point decline in return on average tangible common shareholders’ equity (ROTCE); BPI finds an 80-basis-point decline in ROTCE. See id.

[8] See id.

[9] See 88 Fed. Reg. at 64,564; 64,574

[10] See Regulatory Capital Rule: Large Banking Organizations and Banking Organizations With Significant Trading Activity, 88 Fed. Reg. 64,028, 64,168 (Sept. 18, 2023). The long-term debt proposal itself notes that the proposed capital changes would “lead mechanically to increased requirements for [long-term debt] under the [long-term debt] proposal.”

[11] See 88 Fed. Reg.at 64,564; 64,574

[12] See Total Loss-Absorbing Capacity, Long-Term Debt, and Clean Holding Company Requirements for Systemically Important U.S. Bank Holding Companies and Intermediate Holding Companies of Systemically Important Foreign Banking Organizations, 82 Fed. Reg. 8,266 (Jan. 24, 2017).

[13] The agencies ultimately require a calibration of 6% of RWAS by assuming (i) a capital starting point of zero capital; (ii) an ending point of common equity tier 1 capital equal to 4.5% of RWAs assets plus a stress capital buffer equal to at least 2.5% of RWAs; and (iii) minus a balance sheet depletion allowance of 1%. See 88 Fed. Reg. at 64,530. A 2% alternative calibration more sensibly assumes (i) a capital starting point of 2%; (ii) an ending point of common equity tier 1 capital equal to 4.5% of RWAs without a stress capital buffer (given that the bank will not typically continue to operate); and (iii) minus a (more conservative) balance sheet depletion allowance of 0.5%.

[14] See 88 Fed. Reg at 64,538–39.

[15] See generally, Haelim Anderson, Francisco Covas and Felipe Rosa, The Long-Term Debt Shortfall and the Liquidity Coverage Ratio, Bank Policy Institute (Oct. 23, 2023) (link); See also Anderson et. al, supra note 3.

[16] See Anderson et. al, supra note 3.

[17] See 12 U.S.C. 5365(a)(1); 12 U.S.C. 5365(a)(2)(A).

[18] See 82 Fed. Reg. at 8,267 (“The Board is issuing the final rule under section 165 of the Dodd-Frank Act.”).

[19] See 12 U.S.C. 5365(a)(2)(C).

[20] See 88 Fed. Reg. at 64,534.

[21] Id. at 64,537.

[22] See generally, Comment Re: Long-term Debt Requirements for Large Bank Holding Companies, Certain Intermediate Holding Companies Foreign Banking Organizations, and Large Insured Depository Institutions, Breckinridge Capital Advisors 2 (Sep. 26, 2023) (link) (noting “the rule is likely to reduce the ability of mid-size institutional investors to directly hold long-term bank debt[.]”).

[23] The analysis occurred over the trading weeks beginning on October 10, 2023, October 16, 2023, October 23, 2023, and October 30, 2023, which should represent normal trading activity given the absence of any extraordinary market shocks or other events that would skew typical trading patterns.

[24] A fixed-income portfolio that uses the U.S. Aggregate Bond Index, which tracks corporate bonds and a broader spectrum of instruments, including U.S. Treasuries, as a benchmark is estimated to need at least $800 million in total assets to make a $400,000 investment in the LTD of a single banking organization. According to Morningstar, 355 of 557 mutual funds and ETFs with a U.S. Aggregate Bond Index benchmark have less than $800 million in assets under management.

[25] Registered funds are only considered to be diversified if they do not hold more than five percent of the fund’s total assets in a single issuer. See 15 U.S.C. 80a-5(b)(1).

[26] See 2023 Bank Failures, Financial Stability Board 7 –9 (Oct. 10, 2023) (link).

[27] Insolvency proceedings would be governed by the U.S. Bankruptcy Code, Title II of the Dodd-Frank Act, or the FDIA.

[28] See generally, BPI Comments on Proposed Rule Requiring Long-Term Debt Maintenance, Bank Policy Institute (Jan. 16, 2024) (link)