United States: Citigroup Emergency Liquidity Program, 2008

As of September 30, 2008, Citi was the second-largest bank in the United States, with more than $2 trillion in assets; its largest banking subsidiary, Citibank, N.A., was the third-largest insured depository institution in the country. Citi was also a major counterparty to many institutions globally and played a central role in payments clearing, investment banking, asset management, and brokerage services worldwide (Fed 2008b).

By November 2008, Citi had significant amounts of commercial paper and long-term senior and subordinated debt outstanding. Citi had posted losses of $10.4 billion in 2008 through the third quarter—partially as a result of its exposure to mortgage-related assets. In the weeks leading up to November 23, investors had become concerned about Citi’s viability, and the firm began struggling to raise funding (Fed 2008b). Then–Treasury Secretary Hank Paulson thought that Citi was “teetering on the brink of failure” by mid-November 2008 (SIGTARP 2011, 3).

On November 23, 2008, the Fed said that it would provide contingent financing for certain losses on the to-be-determined asset pool to restore confidence in Citi and support financial stability (Fed 2008b). The purpose of the Asset Guarantee Program, which the Fed’s loan facility supported, was to address market fears that Citigroup faced further catastrophic losses. Officials felt that they had no other options to do this:

In Treasury’s view, asset guarantees would ‘calm market fears about really large losses,’ thereby encouraging investors to keep funds in Citigroup and Bank of America. When asked to discuss possible alternatives to asset guarantees and why they were not selected, Treasury indicated that no alternatives were seriously considered. (COP 2009, 38)

Fed officials believed that a Citi failure would have been destabilizing to the global financial system by disrupting numerous markets. Then–Fed Chairman Ben Bernanke said that a Citi failure could have resulted in blocked access to ATMs and halted the issuance of paychecks. An FDIC official told the Special Inspector General for the Troubled Asset Relief Program (SIGTARP) that the impacts of a Citi failure on money market liquidity could have been global in scope (SIGTARP 2011). The Treasury said that a driving motivation behind the AGP for Citi (and for Bank of America) was a fear that if Citi failed, it would result in “the same deep, systemic damage as Lehman Brothers’ collapse” (COP 2009, 37). On the weekend of November 22–23, 2008, at least one participant in a meeting of officials from the Fed, OCC, and FDIC said the goal of supporting Citi was to avoid a “large worldwide bank run” (SIGTARP 2011, 15).

On Saturday, November 22, 2008, at 3:36am, Citi delivered to the FRBNY a proposal for government assistance. In its proposal, Citi asked that the government cover 100% of losses on a ring-fenced asset pool of $306 billion, for which Citi would compensate the government with $20 billion in preferred shares. Then–Citi CEO Vikram Pandit said the proposal was based off a similar proposal that Citi and the FDIC had crafted for its earlier proposed acquisition of Wachovia (see Appendix). The Citi officials also said it was important to them to get regulators to approve a 20% risk weight for the assets in the asset pool. The lower risk weight would result in a significant reduction in the bank’s regulatory capital requirement and make it easier for it to report to the market that it was in compliance with regulatory requirements. At noon, the government parties rejected the proposal and submitted to Citi a new proposal on Sunday (SIGTARP 2011). The Fed Board of Governors discussed and approved the proposed loan facility at its November 23, 2008, meeting (Fed 2008d). Figure 4 shows the difference between the original Citi proposal (November 22) and the ultimate government proposal (November 23).

Source: SIGTARP 2011.

The loan facility facilitated the Asset Guarantee Program, which also included loss-sharing provisions from the Treasury and FDIC. Citi also received an additional Treasury TARP capital injection (Fed 2008b).

Under the Asset Guarantee Program, the Treasury and FDIC agreed to guarantee ring-fence-eligible assets by taking second- and third-loss positions, respectively, behind a first-loss layer retained by Citi. In compensation for their loss protection, Citi would issue the Treasury and FDIC $4 billion and $3 billion, respectively, in preferred stock bearing an 8% coupon (Fed 2008b). Treasury also received warrants for the purchase of common stock at a strike price of $10.61 per share, representing an exercise value of $705.6 million. That is, regardless of the market price, the government could acquire common stock at $10.61 per share, which would provide $705.6 million of capital to Citi (the government also had the option of selling the warrants without exercising them).

At the same time as it agreed to the AGP, Treasury agreed to purchase $20 billion in senior preferred stock bearing an 8% coupon with TARP funds through Targeted Investment Program (TIP) (see Hoffner and Arnold [2024]). Citi received the TIP capital injection on December 31, 2008 (SIGTARP 2011). As part of TIP, the Treasury received additional warrants for the purchase of common stock at a strike price of $10.61 per share, for an exercise value of $2 billion. In this case, full exercise of the warrants would have provided $2 billion of capital to Citi (Fed 2008b).

By the time it received this government support, Citi had utilized several other broad-based crisis response programs. As of Friday, November 21, 2008, Citi had $24.3 billion outstanding under the Fed’s collateralized liquidity facilities; had $200 million outstanding under the Fed’s Commercial Paper Funding Facility; had borrowed $84 billion from the Federal Home Loan Bank System; and had received $25 billion in TARP capital injections under the Capital Purchase Program (FCIC 2011). Citi’s bank subsidiaries also had standing access to the Fed’s discount window.

On November 22, 2008, at 11:14pm, Citi asked the FRBNY to double its access to the Commercial Paper Funding Facility; the FRBNY denied the request (SIGTARP 2011). On November 23, 2008—concurrent with the announcement of the government package—the Fed authorized the extension of credit to Citi’s London-based broker-dealer under the Primary Dealer Credit Facility (Fed 2009). Citi’s London broker-dealer immediately began borrowing daily from the facility (Fed 2020). By December, Citi had issued $32 billion in senior debt covered by the FDIC’s guarantee program (FCIC 2011). At year-end 2008, Citi also had $60.1 billion in non-interest-bearing deposits in the US, all of which would have likely qualified for the FDIC’s Transaction Account Guarantee Program (Citi 2009b; Vergara 2022). In this program, for which all banks were automatically enrolled unless they opted out, the FDIC provided unlimited deposit insurance, for a fee, to non-interest-bearing transaction accounts and other low-interest-bearing accounts (Davison 2019).

Regulatory Treatment/Relief

As of December 31, 2008, regulators assigned the asset pool a favorable 20% risk weighting for purposes of calculating Tier 1 capital, the standard risk weighting for AAA-rated corporate credits under Basel capital standards (BIS 2004; Citi 2008). A 20% risk weight would have required them to hold just $4.8 billion of capital against the $300.8 billion asset pool. (Under regulatory capital standards, the risk weighting is multiplied by a standard 8% capital requirement, resulting in 1.6% requirement for assets with a 20% risk weighting.) In contrast, the company held roughly $21 billion in capital against the asset pool assets before the transaction, based on its statement that the transaction freed up $16 billion in capital.^FThat $21 billion in capital would represent an average risk weighting of 87%, which would be more appropriate given the composition of the asset pool. Citi later reported in its 2008 annual report (Form 10-K) that the AGP’s lower risk weight had resulted in a 150-basis-point (bp) increase in Citi’s Tier 1 capital ratio at December 31, 2008, calculated by Citi in accordance with normal risk-based capital guidelines (Citi 2009b). Citi reported the following year that exiting the program had resulted in a 157 bp decrease in its Tier 1 ratio. Those figures would suggest the company’s initial risk weighting of the asset pool assets was closer to 70%. According to then–Citi Vice Chairman Ned Kelly, the capital relief obtained by Citi in part through the risk-weighting of the asset pool (and in part through the capital injection) was the reason that Citi ultimately accepted the deal (SIGTARP 2011).

Regulators granted a 20% risk weight to the asset pool, despite Citi’s remaining exposed to the first $39.5 billion in losses. Citi had already reserved $9.5 billion against the assets in the asset pool; the asset pool would ultimately lose $10.2 billion by the time of termination of the AGP (it remains unclear if that $10.2 billion in losses included the original $9.5 billion in loss reserves or was in excess of it; no sources indicate that it was in excess of the loss reserves) (SIGTARP 2012).

Section 13(3) of the Federal Reserve Act (FRA) (12 U.S.C. § 343) provided the legal authority for the Fed to create and lend through the loan facility. The Fed Board voted to invoke Section 13(3) on November 23, 2008 (Fed 2008b). Section 13(3) permitted the Fed, “in unusual and exigent circumstances,” to provide liquidity to “any individual, partnership, or corporation,” provided the assistance was “secured to the satisfaction” of the Fed (US Congress 2008a, vol. 12 U.S.C. Ch. 3 [12 U.S.C. § 343]). The invocation of Section 13(3) allowed the Fed to provide liquidity more broadly than its monetary policy and discount window authorities allowed (US Congress 2008a, vol. 12 U.S.C. Ch. 3, secs. 10B, 13, 14).

Fed legal staff and policymakers had consistently interpreted Section 13(3)’s “secured to the satisfaction” requirement as meaning that the Fed had to be reasonably confident of full repayment (Alvarez 2022). In other words, while the Fed might take some losses in 13(3) lending, to provide liquidity under Section 13(3), the Fed had to make the ex ante determination that it would be repaid (Alvarez 2022). The FRBNY engaged an external vendor to conduct a stress scenario on the asset pool, the conclusion of which was that losses on the asset pool were unlikely to exceed $56.2 billion (the threshold at which the FRBNY would be required to provide a loan) (GAO 2011) (see Key Design Decision No. 10, Balance Sheet Protection).

Yet, notably, the Fed loan would be made only in the event that losses exceeded the protection layers that were subordinated to the Fed. The Fed then–General Counsel Scott Alvarez later said of this that even in that setting of deteriorating collateral values and no recourse (except to interest and Citi’s portion of the loss-sharing), the 90-10 loss-sharing agreements were adequate to secure the Fed to its satisfaction given the greater discount to par value that would then be prevailing and the Fed’s ability to hold the collateral until it recovered value (Alvarez 2022).

Given that the Citi AGP had a maturity of five or 10 years (see Key Design Decision No. 9, Loan Duration), it was possible that the Fed’s loan facility lending commitment could have been active after the “unusual and exigent circumstances” required by Section 13(3) had passed. The Fed’s view on this, according to Alvarez, was that if the ring-fenced losses had penetrated through the junior loss-protection layers to the point where the Fed had to lend into the ring-fence, that situation itself would satisfy the “unusual and exigent circumstances” requirement (Alvarez 2022).

The FRBNY administered the loan facility (Citi, Treasury, FDIC, and FRBNY 2009a). The Fed worked closely with the Treasury and the FDIC on the Citi ring-fence. On January 15, 2009, Treasury Secretary Paulson wrote a public letter to Fed Chair Bernanke saying that the Treasury fully supported the Fed’s participation in the AGP (Paulson 2009).

The specifics of the loan facility were codified in the Citi Master Agreement contract between Citi and the government, signed on January 15, 2009. The Master Agreement also included addenda that governed executive compensation, corporate governance (see Key Design Decision No. 12, Other Conditions), and assets eligible for the ring-fence (Citi, Treasury, FDIC, and FRBNY 2009a). On November 30, 2009, Citi and the government amended the Master Agreement, making technical revisions and including provisions related to the Making Home Affordable program and other loan modification programs (Citi, Treasury, FDIC, and FRBNY 2009b).

Between November 23, 2008, when the government and Citi announced the support package, and January 15, 2009, when the government and Citi signed the Master Agreement, the government and Citi negotiated on the assets that would comprise the asset pool (SIGTARP 2011).

To activate access to the loan facility, the Treasury and FDIC loss-sharing arrangements must have been exhausted (Fed 2008b). In total, losses on the asset pool would have to reach $56.2 billion^FInitial government estimates put the expected 10-year losses at between $34.6 billion and $43.9 billion (SIGTARP 2011). for the FRBNY loan facility funding commitment to be triggered (Fed 2010). Eligible losses under the loan facility included charge-offs and realized losses on collection, short-sale losses, foreclosures, and losses through a permitted disposition or exchange but did not include mark-to-market unrealized losses or an increase in related loss reserves (Fed 2009).

To obtain a loan under the loan facility, the FRBNY would require Citi to submit a borrowing request 20 calendar days before funding. After receiving the borrowing request, the FRBNY would have between 20 and 30 business days to provide the proceeds of the loan to Citi (Citi, Treasury, FDIC, and FRBNY 2009a).

In the event that the asset pool incurred losses during a time in which the loan was extended, Citi would be required by the Master Agreement to prepay a principal amount, with accrued interest, on the loan facility equivalent to 10% of that loss value within 30 days after the end of the calendar quarter in which such losses were incurred (Citi, Treasury, FDIC, and FRBNY 2009a).

Citi also would have paid interest on the loan if the loan had been activated. Given that the loan was a one-time loan for the full amount (in other words, there could be no multiple draws), interest would be paid on that total amount. The amount of that loan would be 90% of the remaining Fed-covered asset pool (in other words, the total Fed coverage less the 10% prepayment). Failure of Citi to pay interest on the loan within five days of its coming due would be an event of default under the Master Agreement (Citi, Treasury, FDIC, and FRBNY 2009a). Citi managed all the assets in the asset pool through Citi Financial—Citi’s US consumer finance business subsidiary (Citi 2009a).

The FRBNY contracted two outside vendors, without a competitive bidding process, to perform valuation services in connection with the loan facility: PricewaterhouseCoopers, contracted December 1, 2008, for a fee of $7.8 million; and BlackRock, contracted on December 14, 2008, for a fee if $12.7 million (GAO 2011).

Section 129 of the Emergency Economic Stabilization Act of 2008 (EESA), passed on October 3, 2008, required the Fed to report to the Senate Committee on Banking, Housing, and Urban Affairs (Senate Banking Committee) and the House Committee on Financial Services (House Finance Committee) on any use of its Section 13 authority within seven days of invoking that authority (see Key Design Decision No. 6, Communication and Disclosure) (US Congress 2008b, sec. 129[a]).

On January 13, 2011, SIGTARP issued a 77-page audit report entitled “Extraordinary Financial Assistance Provided to Citigroup, Inc.” (SIGTARP 2011). SIGTARP ultimately found that “the Government constructed a plan that not only achieved the primary goal of restoring market confidence in Citigroup, but also carefully controlled the overall risk of Government loss on the asset guarantee” (SIGTARP 2011, 41). Nonetheless, although SIGTARP found that the government parties had effectively avoided financial catastrophe and protected taxpayer funds, it also found that the government’s conclusions with respect to Citi were “strikingly ad hoc” and “undoubtedly” contributed to moral hazard (SIGTARP 2011, 41).

In July 2011, the GAO published a report on the Fed’s emergency programs, including an appendix specifically on the Citi loan facility, and found that the FRBNY did not use a competitive procurement process to award contracts to the private sector entities hired to provide services (see Key Design Decision No. 4, Administration) (GAO 2011).

On November 6, 2009, the COP published its report on US government guarantees during the Global Financial Crisis of 2007–2009. The oversight panel said that government guarantees in general had advantages for taxpayers—such as bearing no up-front price tag and resulting in profits to the government—but that the guarantees resulted in moral hazard and market distortions. The COP said that one possible reason other AGPs were not extended by the government was that the announcements of the two AGPs (for Citi and Bank of America) sufficiently calmed the market. The panel said that “the asymmetric nature” of some of the terms of the Master Agreement protected the government and disadvantaged Citi, lowering the likelihood that the government would have to pay out on the guarantee (COP 2009, 41). The COP also said that the rationale for the Citi AGP remained unclear and that more transparency would have assisted in evaluating and overseeing the program. Citi and Treasury reported to the Oversight Panel “substantial” monitoring and auditing of the asset pool (COP 2009, 40). The Oversight Panel said that it would itself closely monitor the performance of the asset pool (COP 2009).

The government aimed to calm markets by communicating a “decisive action” before Asian markets opened on Sunday, Eastern US time (SIGTARP 2011, 19). On a conference call on Thursday, November 20, 2008, FRBNY President Timothy Geithner told Chairman Bernanke, Secretary Paulson, FDIC Chairman Sheila Bair, and OCC Comptroller John Dugan, “We’ve told the world we’re not going to let any of our major institutions fail. We are going to have to make it really clear we’re standing behind Citigroup” (SIGTARP 2011, 13). In an email exchange on November 22, 2008, an email between senior FDIC officials said, “the main point is to let the world know that we will not pull a Lehman,” to which the recipient responded, “At this stage, it is probably appropriate to be clear and direct that the US government will not allow Citi to fail to meet its obligations” (FCIC 2011, 380). Jason Cave, an FDIC official party to the AGP government negotiations, said that the Fed’s participation (via its contingent loan facility) primarily added value through its announcement effect, signaling government unity in standing behind Citi (Cave 2024).

On Sunday, November 23, 2008, the government announced the AGP for Citi. The government said that its actions to support Citi supported its commitment to financial market stability. The government said that it would strengthen the financial system and protect taxpayers and the broader US economy (USG 2008). Regulators thought that the announcement of support to Citi needed to be made over the weekend to support market confidence before the market open on Monday (FCIC 2011).

In Citi’s November 24, 2008, press release announcing the transaction, Citi emphasized the regulatory capital benefit of the Asset Guarantee Program. The headline of the press release was, “Citi Adds $40 Billion of Capital Benefit through Agreement with U.S. Treasury, Federal Reserve, and FDIC” (Citi 2008).^FThe $40 billion figure included $16 billion in capital reduction due to the lower risk weighting of 20% that regulators had agreed to allow Citigroup to apply to the $306 billion asset pool in its calculation of regulatory capital (see Key Design Decision No. 2, Part of a Package). The total also included half of the $7 billion in preferred stock that Citi agreed to issue to the FDIC and Treasury as a fee for the Asset Guarantee, because regulators agreed to include only half that amount as capital. The remaining $20 billion was the TARP preferred stock from the Treasury (Citi 2008).

Citi CEO Vikram Pandit said during Citi’s Q4 2008 earnings call that the AGP helped ensure that Citi was well capitalized. He added that Citi viewed it—along with the TIP capital injection—as a bridge to rebuilding tangible common equity, which he said “there are some concerns about” (Citi 2009a). He said that the asset guarantee had helped Citi to reduce its risky assets. Citi made specific reporting on the asset pool available to its investors and the government (Citi 2009a).

Section 129 of the EESA required the Fed to report to the Senate Banking Committee and House Finance Committee on any use of its Section 13(3) authority within seven days of invoking that authority, including its justification for invoking Section 13(3) and the terms of any Fed actions, specifically, “the size and duration of the lending, available information concerning the value of any collateral held with respect to such a loan, the recipient of warrants or any other potential equity in exchange for the loan, and any expected cost to the taxpayers” (US Congress 2008b, sec. 129[a][2]). Pursuant to its Section 129 reporting requirements, the Fed issued a Section 129 report on its loan facility on November 23, 2008. In its Section 129 Disclosure, the Fed said that the AGP—of which its loan facility was part—would “help restore confidence” in Citi and would “promote financial stability” (Fed 2008b, 2). The Fed also said that it did not expect its loan facility to result in any losses to the Fed or the taxpayer and presented other disclosures required by Section 129 (subsection [a][2]) (Fed 2008b).

On January 15, 2009, the day on which Citi and the government signed a Master Agreement defining terms of the AGP, Secretary Paulson wrote a public letter to Chairman Bernanke saying that the Treasury understood that the FRBNY’s participation was necessary, that the Treasury fully supported the Fed’s actions, and that the Treasury acknowledged that it could receive lowered remittances from the Fed to the Treasury’s general fund as a result of possible losses on the facility (Paulson 2009).

The Fed discussed the loan facility, the motivation for the AGP, the timing of the interventions, and the outcomes of the interventions in its 2008 annual report. In that report, the Fed said that the FRBNY loan commitment’s probable loss and fair value were both zero because it was unlikely that the FRBNY would have to make the loan (Fed 2009). Citi also disclosed the loan facility in its 2008 annual report (Citi 2009b). The Fed described the loan facility termination in its 2009 annual report (Fed 2010). Citi also disclosed the termination of the loan facility in its 2009 annual report (Citi 2010).

COP member Damon Silvers said that Citi did not disclose to the COP a comprehensive and itemized list of assets under the asset pool, thereby rendering the COP unable to independently value the Citi ring-fence (COP 2009).

According to the original term sheet, the size of the overall asset pool subject to the ring-fence arrangement was $306 billion (Fed 2008a). Then–Citi Vice Chairman Kelly told the FCIC that “there was not a huge amount of science in coming to that [$306 billion] number” and that the figure had been chosen by regulators to “give the market comfort that the catastrophic risk has been taken off the table” (FCIC 2011, 381). When Citi and the government signed the Master Agreement for the AGP in January 2009, the final agreed size of the asset pool was capped at $300.8 billion (Citi, Treasury, FDIC, and FRBNY 2009a). In its 2008 annual report, the Fed reported its share as a contractually committed amount of $244.8 billion (Fed 2009). The Fed likely would have funded this liquidity provision through the creation of new reserves.

In the event total losses surpassed the $56.2 billion threshold, the FRBNY agreed to make a loan equal to:

(a) the sum of the Adjusted Baseline Values of all Covered Assets as of the end of the most recently completed Calendar Quarter prior to the FRBNY Funding Date less (b) the principal amount of the FRBNY Loan that would otherwise be subject to immediate prepayment by Citigroup pursuant to Section 4.7(a), which amount (b) shall be calculated as 10% of the excess of (i) the Citigroup Quarterly Net Loss corresponding to the Covered Loss giving rise to the funding of the FRBNY Loan over (ii) the dollar amount equal to (A) the amount of such Covered Loss funded by any Treasury Advance and/or FDIC Advance divided by (B) 0.90. (Citi, Treasury, FDIC, and FRBNY 2009a, 7–8)

Figure 5 shows the mechanics of the $56.2 billion threshold for the Fed’s lending obligation.

Source: SIGTARP 2011.

The Fed charged a floating rate of the three-month Overnight Index Swap (OIS) rate plus 300 basis points (bps) on the loan (which would be equal in size to 90% of the remaining assets in the asset pool) (Citi, Treasury, FDIC, and FRBNY 2009a).

The OIS rate is a market-determined rate representing expectations for the Fed’s policy rate over different tenors; in this case, three months. At the time, the Fed’s “Regulation A”—regulations written by the Fed to guide its implementation of Section 13(3), among other statutes—called for Fed emergency credit to carry an interest rate “above the highest rate in effect for advances to depository institutions,” a reference to the Fed’s discount window (Fed 2008c, vol. Code of Federal Regulations, title 12, sec. 201.4). At the time, the discount window’s primary credit was 25 bps above the upper limit of the Fed’s policy rate target range, for a sum total of 50 bps; the rarely used “secondary credit” charged the primary credit rate plus 50 bps (100 bps total) (Fed 2008e). According to the Congressional Oversight Panel, the interest rate was “standard and within commercial limits” (COP 2009, 41).

For any amounts overdue, the interest rate would be the “default rate”—the normal interest rate (OIS + 300 bps) plus an additional 200 bps (Citi, Treasury, FDIC, and FRBNY 2009a).

Upon termination of the Master Agreement on December 23, 2009, Citi paid the FRBNY a $50 million termination fee and agreed to reimburse the FRBNY for its out-of-pocket expenses associated with administering the loan facility (Citi, Treasury, FDIC, and FRBNY 2009c; Fed 2010).

The loan would be a term loan, with final payments due no later than November 20, 2018. No amounts repaid or prepaid could be reborrowed by Citi (Citi, Treasury, FDIC, and FRBNY 2009a). Citi could voluntarily prepay the loan without penalty (Citi, Treasury, FDIC, and FRBNY 2009a).

Citi would be required by the Master Agreement to repay all debts (inclusive of interest) on any loan under the loan facility in two payments. The Master Agreement required Citi to pay the “Installment Balance” by November 20, 2013—a five-year maturity—and the remaining balance of the loan by November 20, 2018—a 10-year maturity—which the FRBNY could extend unilaterally by one year. The Installment Balance would be equal to the share of the available amount (see Key Design Decision No. 7, Source and Size of Funding) represented by nonresidential assets, less Citi’s 10% loss sharing value, less 90% of proceeds earned on any asset in the asset pool (Citi, Treasury, FDIC, and FRBNY 2009a).

From November 21, 2008, through December 31, 2008, Citi took $900 million in losses on the asset pool (Citi 2009b). As of June 30, 2009, Citi had announced a cumulative $5.3 billion in losses on the asset pool (COP 2009). On December 13, 2009, Citi informed the Fed that it intended to terminate the Master Agreement. The next day, the Fed granted approval for the Citi-proposed termination, subject to the condition that Citi issue common stock and other securities and repay its TIP capital.^FThe other securities comprised common stock overallotment options, tangible equity units, and employee stock options (Hoffner and Arnold 2024). On December 22, 2009, Citi raised equity and obtained the Treasury’s approval to repay its TIP investment, allowing Citi to exit the AGP.^FCiti raised equity through the issuance of common and preferred stock, as well as other securities but did not satisfy the overallotment options minimum requested by the government. While the government allowed Citi to exit the AGP that day, Citi later raised more equity in March 2010 through the issuance of trust preferred securities (Hoffner and Arnold 2024). The following day, Citi and the government terminated the Master Agreement,^FOn the same day, Citi also repaid $20 billion in TIP preferred shares (see Hoffner and Arnold [2024]) (Citi 2010). thereby terminating the FRBNY’s loan facility (Citi, Treasury, FDIC, and FRBNY 2009c).^FAlso on that date, the Treasury agreed to reduce Citi’s liability for the asset guarantee–linked trust preferred securities from $4 billion to $2.2 billion; including the securities issued to the FDIC, the government continued to hold $5.3 billion of Citi preferred securities at the end of 2009 (the Treasury and FDIC had converted their preferred shares into trust preferred securities in 2008, increasing their loss-absorbing capacity). Citigroup also repaid the Treasury for the $20 billion of TARP trust preferred securities on December 23, 2009 (Citi 2009b). 

The FRBNY never expected to lend through the loan facility. For the Fed’s loan facility to be triggered, losses on the asset pool would have had to exceed $56.2 billion, but even in the highest-stress modeled scenario, the government expected that losses would not exceed $43.9 billion (ultimately, the government put expected losses at $37 billion). Later, in May 2009, BlackRock estimated a base-case scenario of $32.7 billion in losses and a stress scenario of $50.8 billion (SIGTARP 2011). Ultimately, Citi’s losses on the asset pool were just $10.2 billion (SIGTARP 2012).

While ultimately Citi and the government agreed to decrease the size of the asset pool by $6.5 billion, it was not a symmetric reduction: they reduced corporate securities, loans, and lending commitments by $47.3 billion, while increasing consumer loans and lending commitments by $40.8 billion (SIGTARP 2011). The ultimate asset pool consisted mainly of residential and commercial mortgage-backed securities and loans (FCIC 2011).

To be eligible for inclusion in the asset pool, the assets had to satisfy the following criteria:

• Be included on the balance sheet of a Citi affiliate by November 21, 2008;
• Not be a foreign asset;
• Not be an equity or derivative asset;
• Be issued or originated before March 14, 2008;
• Not have Citi as an obligor; and
• Not be guaranteed by the government under another agreement.

Eventually, Citi replaced $2.3 billion of assets that the government rejected for inclusion in the asset pool (SIGTARP 2011).

Figure 7 shows the exclusions and substitutions made to the asset pool before it was finalized.

Source: SIGTARP 2011.

The FRBNY would have exclusive, first-priority perfected security interests in each covered asset, which Citi would pledge as collateral for the Loan (Citi, Treasury, FDIC, and FRBNY 2009a).

When Citi reported its gains and losses on the asset pool to the Treasury (see Key Design Decision No. 12, Other Conditions), Citi would be required to remit a certain share of any gains to the government (COP 2009). In the event that either the FRBNY or Citi made gains on covered assets during the period the loan was extended, Citi would be required to pay 90% of those gains toward the principal of the loan, thereby reducing its size. In the event that Citi failed to use gains or recoveries on covered assets to repay the principal on the loan as stipulated in Section 7 of the Master Agreement, it would be an event of default (Citi, Treasury, FDIC, and FRBNY 2009a). As assets in the asset pool amortized or were sold, the overall size of the asset pool—and thereby the Fed’s contingent lending obligation—would diminish over time (COP 2009).

Any loans through the loan facility would be made with only partial recourse to Citi for principal—to the extent that Citi had an ongoing 10% loss-sharing arrangement with the Fed—but would be made with recourse to interest payments (Fed 2008b). Full-recourse lending is standard practice for the Fed’s discount window, though some of its Section 13(3) lending has been done on a nonrecourse basis (Fed 2021; Kelly 2023).

Our research did not uncover any specific impact of the loan facility on monetary policy transmission. No lending ultimately occurred under the loan facility.

As conditions of the overall AGP, the government required Citi to report on a quarterly basis: (a) the adjusted baseline value of each asset; (b) aggregate losses incurred on the asset pool by asset class; and (c) aggregate gains or losses on the asset pool. The government also required Citi to produce monthly reports to the Treasury including updates on the value of assets (reported as fair value, not mark-to-market) in the asset pool, the monthly changes of the asset values, the year-to-date changes in the value of those assets, the main determinants of the changes in those values, and Citi’s forecasted lifetime losses or gains on those assets based on Citi’s stress testing (COP 2009).

As part of the AGP broadly (inclusive of the loan facility), Citi was required by the Master Agreement to abide by numerous conditions regarding, inter alia, asset management, corporate management, executive compensation, and dividend payments (capped at one cent per share of common stock). Additionally, the Treasury TARP capital injections (see Key Design Decision No. 2, Part of a Package) came with their own conditions, including compliance with executive compensation and foreclosure-mitigation policies (Fed 2008b). On November 30, 2009, Citi and the government revised the Executive Compensation Guidelines associated with the Master Agreement and backdated that revision to July 23, 2009 (Citi, Treasury, FDIC, and FRBNY 2009a; Citi, Treasury, FDIC, and FRBNY 2009b). Figure 8 shows the Governance and Asset Management Guidelines, Exhibit B to the Master Agreement.

Source: SIGTARP 2011.