The United States’ Debt Guarantee Program of the Temporary Liquidity Guarantee Program

In addition to the DGP, the TLGP had a second component called the Transaction Account Guarantee Program that provided a blanket guarantee on non-interest-bearing transaction accounts at participating banks and thrifts. These transaction accounts were of the type commonly used to meet payroll and other business purposes. The TAGP sought to prevent a run on such accounts by providing them with a full guarantee. At introduction the TAGP was designed to last until December 31, 2009, but the FDIC subsequently extended this termination date to December 31, 2010.

U.S. policymakers also introduced (or first explained) two additional programs on October 14^th that were seen as linked to the DGP. The Capital Purchase Program (CPP) announced by Treasury at the same press conference as the TLGP provided up to $250 billion for capital injections into viable financial institutions. Authorities informed the nine major financial institutions that were the initial recipients of capital injections pursuant to the CPP that participation in the CPP was a requirement for accessing the DGP. This was both (a) so that the institutions would have larger capital buffers and thus be less likely to impose losses on the DGP by defaulting on guaranteed debt and (b) to avoid the stigma that might result from certain major financial institutions participating in the DGP while others declined to do so. No other institutions were explicitly required to accept capital via the CPP in order to participate in the DGP, but the two programs were generally seen as complementary.

On October 14, the Federal Reserve Board also provided details of the Commercial Paper Funding Facility (CPFF) it had enacted on October 7. Under the CPFF, a specially created limited liability company would use funds borrowed from the Federal Reserve Board of New York to purchase eligible commercial paper directly from issuers. This was intended to provide liquidity to commercial paper issuers otherwise unable to access it due to strains in the commercial paper market.    

The FDIC established the TLGP (and thus the DGP) under the authority of the FDIC Improvement Act of 1991 (FDICIA). Enacted in the wake of a series of thrift failures in the late 1980s and early 1990s that left the FDIC’s insurance fund seriously undercapitalized, the FDICIA required “least-cost resolution,” meaning that the FDIC had to select a method of providing assistance to troubled insured depository institutions that would minimize the cost to the Deposit Insurance Fund. However, the law included a “systemic risk” exception that authorized the FDIC to bypass the least-cost resolution requirement if following it “would have serious adverse effects on economic conditions or financial stability” and overlooking it “would avoid or mitigate such adverse effects.” In the event of any loss to the Deposit Insurance Fund stemming from an action undertaken based on the systemic risk exception, the law requires the FDIC to recover that loss through “1 or more special assessments on insured depository institutions, depository institution holding companies (with the concurrence of the Secretary of the Treasury with respect to holding companies), or both, as the [FDIC] determines to be appropriate.”^F12 U.S.C. 1823(c)(4)(G)

On October 13, 2008, the Secretary of the Treasury, in consultation with the President, invoked the systemic risk exception to facilitate the TLGP based on the recommendations of at least two-thirds of the FDIC’s Board of Directors and two-thirds of the Federal Reserve Board of Governors, all as required by law.^FAfter the crisis, the FDIC’s use of the systemic risk exception to enact a widely available guarantee program came under significant legal scrutiny, and the Dodd-Frank Act curtailed the FDIC’s authority to introduce a program like the DGP.

Program documents did not contain any mention of a limit on the amount of guaranteed debt that could be issued under the DGP.

While the FDIC was working with the Treasury and Federal Reserve on the development of the DGP, an initial difference of opinion existed on the range of institutions that should be eligible for the program. The FDIC, whose historical mandate and experience revolved around the protection of insured depository institutions, sought to limit the program to such institutions. The Treasury and Federal Reserve argued for the inclusion of a broad range of nonbank financial institutions. Ultimately, those eligible for the DGP included insured depository institutions and bank and financial holding companies. The inclusion of holding companies stemmed from the observation that the senior unsecured debt covered by the DGP was typically issued at the holding company level in most holding company structures, with the holding companies then providing liquidity to their depository institution subsidiaries. (Federal Register 73 [229]: 72244 and 72250, 2008)

The DGP also allowed for the participation of other affiliates of insured depository institutions, provided that such affiliates would be admitted on a case-by-case basis in the sole discretion of the FDIC based on “such factors as (1) the extent of the financial activity of the entities within the holding company structure; (2) the strength, from a ratings perspective, of the issuer of the obligations that will be guaranteed; and (3) the size and extent of the activities of the organization” (Ibid.).

The rules implementing the DGP enabled the FDIC, working with the other federal regulators, to monitor and limit use of the facility by weaker institutions. In particular, institutions with the lowest supervisory ratings were barred from participating.

The FDIC designed the DGP’s enrollment structure to convince as many eligible institutions as possible to participate, regardless of their credit risk (Technical Briefing on the TLGP 2, 2008). The FDIC believed that automatic enrollment and an opt-out structure would normalize program participation, encourage stable banks to issue guaranteed debt, and collect sufficient fees to cover potential program losses. Moreover, the FDIC played on institutions’ risk aversion by not allowing them to rejoin the program later if they opted out by the December 5 deadline (initially November 12, but later extended). Some institutions without financing problems in October and November still chose to participate, since they worried they would suffer unexpected illiquidity in the future.

Officials wanted to maximize participation in the program for two reasons. First, they believed that strained access to credit was widespread, so many firms would have to participate to meaningfully calm markets. Second, they wanted to minimize adverse selection, where firms most likely to default would be the only ones participating. If the market viewed participation as a signal that a firm was insolvent, then the DGP would not cut borrowing costs for solvent firms with liquidity problems. Additionally, if only troubled firms participated, then the FDIC would have to either (1) suffer massive losses as a result of the program; or (2) increase participation fees, which would only exacerbate adverse selection and increase borrowing costs further (GAO, 2009).

The one exception to the DGP’s opt-out structure were the insured depository institution-affiliates discussed above, who had to apply to the FDIC to participate.

The purpose of the DGP was to promote liquidity and stability in interbank unsecured lending markets. To do so, regulators believed it best to only guarantee senior debt.

The DGP specifically excluded unsecured debt with embedded instruments for two reasons. First, the FDIC wanted to promote stability, confidence, and transparency in securities markets, but “not to encourage innovative, exotic or complex funding structures” (Federal Register 73 [229]: 72245, 2008). Second, officials worried that weak banks would use derivatives to obscure underlying risk (Federal Register 73 [229]: 72252, 2008). As such, extending guarantees to those instruments could expose the FDIC to additional risk without meaningfully improving liquidity.

When Treasury Secretary Paulson and Federal Reserve Chair Bernanke approached the FDIC Chair Shelia Bair about guaranteeing bank debt, Bair was reluctant to proceed. She worried that the program would potentially expose the FDIC to huge losses and questioned how such a guarantee program fulfilled the FDIC’s depositor protection role. However, recognizing the risks of a prolonged liquidity crisis, she proposed a program that would guarantee 90% of newly-issued senior unsecured debt. Paulson and Bernanke insisted that a 90% guarantee would not be sufficient to calm markets, and Bair compromised on a program that would cover 100% of certain debt, but only within a limited time frame (Bair, 2011).

Initially, debt of any maturity could be issued pursuant to the DGP. The guarantee on such debt would then last until maturity or June 30, 2012 (whichever came first). Subsequent amendments later altered this approach.

First, the Final Rule issued on November 21, 2008, excluded debt with maturities of 30 days or less. The FDIC observed that, while other federal programs had improved liquidity in short-term funding markets, institutions still struggled to access longer-term unsecured funding. Thus, the DGP would “help institutions to obtain stable, longer-term sources of funding where liquidity is currently most lacking” by focusing on maturities of more than thirty days (Federal Register 73 [229]: 72251, 2008).

Second, an amendment to the DGP adopted on June 3, 2009, extended the life of the guarantee to December 31, 2012, for debt issued on or after April 1, 2009. The FDIC adopted this extension of the guarantee alongside an extension of the issuance window for guaranteed debt from June 30, 2009, to October 31, 2009, as described in more detail below.

During the life of the DGP, some market participants suggested that the FDIC extend the guarantee to as long as ten years given that “real money investors” such as pension funds and insurance companies were more active in longer-term markets. However, the FDIC rejected these suggestions, arguing that extending guarantees to longer-term debt would not be necessary to improve liquidity to interbank and unsecured term debt markets (Federal Register 73 [229]: 72253, 2008).

Program documents did not contain any language restricting the currencies that were eligible for the DGP.

In order to ensure that the DGP was used to rollover existing debt rather than significantly expand debt issuance, the FDIC imposed a cap on the amount of guaranteed debt that a participating institution could issue. The FDIC set this cap at 125% of senior unsecured debt outstanding on September 30, 2008, that would mature by June 30, 2009.

Initially, institutions with no senior unsecured debt outstanding on September 30, 2008, could apply to the FDIC for eligibility to issue an amount of debt to be determined by the FDIC. However, concerns that such an approach might delay determinations of eligibility amounts resulted in the exploration of other options. In the Final Rule issued on November 21, 2008, the FDIC established a cap equal to 2% of consolidated total liabilities as of September 30, 2008, for institutions with no senior unsecured debt outstanding. For all participating institutions, the FDIC could increase or decrease these caps should it deem it necessary after consultation with the appropriate federal banking agency.

These caps were particularly significant because, with one exception noted below, participating institutions were not permitted to issue non-guaranteed debt until the caps were reached. Eligible institutions strongly opposed this measure. Critics made numerous arguments in favor of allowing any institution to issue non-guaranteed debt at will before hitting the cap, including that a prohibition on non-guaranteed issuance would limit flexibility, cause healthy banks to avoid the program, and result in a competitive disadvantage relative to the UK’s Credit Guarantee Scheme (Federal Register 73 [229]: 72255-6, 2008).

Despite these concerns, the FDIC decided to maintain the prohibition on non-guaranteed issuance under the cap for several stated reasons. The “[f]irst, and most important” reason the FDIC retained the rule was to avoid adverse selection, where banks only sought to have guaranteed the debt that they were least likely to repay. On a more pragmatic note, the FDIC worried that widespread issuance of non-guaranteed debt under the cap would lead to confusion about what bonds received the guarantee, while the existing structure would calm markets through consistency (Federal Register 73 [229]: 72255-6, 2008).

The one exception to the prohibition on non-guaranteed issuance was for institutions that affirmatively made an election by a specified date to have the right to issue non-guaranteed debt under the cap and paid a fee for the right equal to 37.5 basis points times the amount of unsecured debt outstanding as of September 30, 2008, maturing by June 30, 2009. Such institutions could issue non-guaranteed debt with maturities greater than June 30, 2012, at any time.

The FDIC also barred banks from using the proceeds from guaranteed debt to prepay debt that was not guaranteed.   

Initially, the FDIC charged a flat annualized rate of 75 basis points on all guaranteed debt, a weighted average of CDS spreads in 2007 (before the subprime crisis hit) and September 2008 (when spreads reached several hundred basis points). The 75-basis-point figure represented the sentiment that by late 2008, default risk was higher than usual but lower than during the height of the crisis (Technical Briefing on the TLGP 1, 2008).

However, the 75-basis-point fee proved prohibitively high for short-term borrowing, with institutions reliant on short-term credit threatening to opt out of the program entirely. In response, in the Final Rule adopted November 21, 2008, the FDIC introduced a range from 50 basis points to 100 basis points based on maturity as follows:

This range was increased with the June 3, 2009, amendment to the DGP extending its issuance window and guarantee. For debt issued on or after April 1, 2009, surcharges of 25 basis points (for insured depositories) or 50 basis points (for other participants) would be applied. For any debt issued pursuant to the Emergency Guarantee Facility, a minimum annualized fee of 300 basis points would apply.

The FDIC considered adopting a fee structure where institutions with a higher credit risk paid higher premiums. Such an approach ensures riskier firms bear higher costs for participating in the program, maintaining market discipline. However, officials said they opted against this approach for practical reasons. The FDIC did not have the time to properly assess differential risk among banks. In addition, the FDIC lacked the regulatory infrastructure and statutory authority to assess the riskiness of bank holding companies and nonbank financial institutions, which, as non-insured institutions, did not fall under the FDIC’s jurisdiction (COP 2009).

Officials at the Federal Reserve urged the FDIC to include bank holding companies and thrifts in the guarantee program as a crucial way to restore liquidity. However, including holding companies and thrifts initially introduced a legal challenge. In the event that program losses exceeded fees, the FDIC had planned to make use of its “special assessment” authority to collect a special fee from all institutions with federal deposit insurance. Bank holding companies and thrifts were not FDIC-insured, meaning the FDIC did not have the statutory authority to impose additional fees on them. To compensate for the fact that they would not be liable in the event of program losses, the FDIC decided to raise the standard assessment fee by 10 basis points on holding companies where FDIC-insured institutions consisted of less than half of the firm’s portfolio (Federal Register 73 [229]: 72251, 2008).

Recognizing this legal challenge, Congress amended FDICIA to allow the FDIC to collect special assessments on depository institution holding companies in May 2009 as part of the Helping Families Save Their Homes Act (GAO, 2010).

The CPP imposed several conditions on participating institutions, including restrictions on executive compensation and a prohibition on increasing dividends. These conditions were not adopted for the DGP.

Initially, the FDIC said it would only fulfill the guarantee if the debtor declared bankruptcy/entered resolution, at which point the FDIC would pay the creditor the remaining balance plus interest and require the creditor to turn over all claims to the debtor’s assets to the FDIC. However, some worried that this approach would prevent the timely exercise of the guarantee and curtail demand for guaranteed debt. In particular, ratings agencies indicated an unwillingness to treat guaranteed debt as government securities (and therefore AAA rated) unless payment was made on default. Thus, if guarantee payments were not delivered in a timely manner, then only the largest institutions deemed already creditworthy would be able to access interbank lending markets, locking out smaller firms without credit ratings. Additionally, since the UK’s Credit Guarantee Scheme paid creditors following the first default on payment, participants worried that UK banks would receive a competitive advantage over U.S. ones (Federal Register 73 [229]: 72263, 2008). For these reasons, the FDIC changed the DGP such that the first payment default on covered debt would trigger the guarantee.

In spring 2009, the FDIC believed that liquidity in financial markets had not returned to pre-crisis levels and worried that an abrupt end to the DGP would lead to market disruption. To ensure an orderly termination of the DGP, officials designed the program’s phase-out to make it slowly uneconomical to issue guaranteed debt (Treasury, 2009). The first extension, from April 1, 2009, to October 31, 2009, raised the participation fee by 25 basis points for FDIC-insured institutions and 50 basis points for holding companies and thrifts. Eligible institutions that had not issued guaranteed debt by that date had to apply for inclusion in the program to prevent program participation from ballooning. All fees collected under the extension were directly deposited into the Deposit Insurance Fund to shore up the FDIC’s balance sheet. The FDIC guaranteed debt issued under the extension through December 31, 2012 (Federal Register 74(105), 2009).

By October, the FDIC recognized that certain firms, due to exogenous market disruptions or other factors beyond their control, found it difficult to roll over their guaranteed debt. Therefore, officials established the Emergency Guarantee Facility to guarantee debt issued between October 31, 2009, and April 30, 2010. Institutions had to apply with the FDIC to issue guaranteed debt until April 30, 2010. The FDIC charged a participation fee of at least 300 basis points, and only firms that previously issued DGP-guaranteed debt were allowed to apply (to limit participation). The FDIC designed the extension to maximize regulatory discretion. All institutions approved to participate in the emergency facility had to be in some extreme, unique circumstance, so FDIC granted regulators the authority to carefully tailor participation requirements on a case-by-case basis (Federal Register 74 [204], 2009). Ultimately, no institutions made use of the Emergency Guarantee Facility.