United States: Citigroup Capital Injection, 2008

According to the Treasury, the purpose of the TIP was to foster financial stability, strengthen the economy, and protect jobs by preventing the loss of confidence in a financial institution and the resulting destabilizing market disruptions (Treasury 2009b).

With respect to eligibility, the Congressional Oversight Panel noted that while the CPP had been a voluntary program for healthy banks, the TIP (and the Systemically Significant Financial Institution [SSFI] Program) addressed institutions “experiencing more difficulty or at risk of failure” in connection to other government programs (COP 2009a, 10, 43). The Treasury would select TIP participants on a “case-by-case basis” and would consider various eligibility criteria including: a bank’s interconnectedness, systemic importance, access to other sources of private or government assistance, size of potential risk for a loss of confidence in relation to illiquid or distressed assets, and whether other institutions with similar vulnerabilities might be affected by a bank’s participation in the program (Treasury 2009b).

According to then–Secretary of the Treasury Paulson and then–President of the FRBNY Geithner, the main purpose of the USG support package was to “assure the world that the Government was not going to let Citigroup fail” (SIGTARP 2011a).

As of September 30, 2008, Citi was one of the largest financial institutions both domestically and in the entire world. It 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 a major supplier of credit to the US and other countries around the world. It had $277 billion and $500 billion in domestic and foreign deposits, respectively. 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).

FDIC staff, in recommending the systemic risk exception, said that a Citi failure “would significantly undermine business and household confidence” (FCIC Report 2011, 381). Regulators were also concerned that a possible Citi failure might lead to economic impacts that could undermine the effects of the Capital Purchase Program under TARP (FCIC Report 2011). Secretary Paulson thought that Citi was “teetering on the brink of failure” by mid-November 2008 (SIGTARP 2011a, 3). At the time, Fed regulators believed that a Citi failure would have been destabilizing to the global financial system by disrupting numerous markets, including those for interbank credit, other credit, derivatives, qualified financial contracts, bank debt, and senior subordinated debt. Then-Chairman Bernanke said that a Citi failure could have resulted in blocked access to automatic teller machines and halted issuance of paychecks. An FDIC official told SIGTARP that a Citi failure’s impacts on money market liquidity could have been global in scope (SIGTARP 2011a). Then–Office of Thrift Supervision Director John Reich said Citi’s situation was “obviously a systemic risk situation”; Secretary Paulson said, “if Citi isn’t systemic, I don’t know what is” (SIGTARP 2011a, 15). On the weekend of November 21–23, 2008, at least one meeting participant (representatives from the Fed, OCC, and FDIC) said the goal of Citi support was to avoid a “large worldwide bank run” (SIGTARP 2011a, 15).

A Fed official said that the Citi preferred-share capital injection was “a clever way for the government parties to provide more protection and be more protected themselves” (SIGTARP 2011a). Ultimately, policymakers judged that the additional exposure to potential losses of $20 billion of TARP funds would be less costly to the financial system than the cost of a Citi failure (SIGTARP 2011a).

Two days after the Citi capital injection (on January 2, 2009), the Treasury said that it had provided the Citi capital injection under a newly named program (when it also named the ring-fencing arrangements ex post as the Asset Guarantee Program) (SIGTARP 2011a). Treasury could have provided the capital injections under the existing Systemically Significant Failing Institution Program, which it had recently created to support the failing insurer American International Group (AIG), but it chose not to do so in part because it did not want to label Citi a failing institution (SIGTARP 2011a).^FThe Treasury announced the Systemically Significant Failing Institution (SSFI) Program in November 2008, “designed to provide stability and to prevent disruption to financial markets from the failure of a systemically significant institution” (Treasury 2008). SSFI support would be provided on a case-by-case basis, and on November 25, 2008, the Treasury made the first disbursement, purchasing $40 billion of senior preferred shares from AIG (Treasury 2008). The SSFI Program was ultimately renamed to the AIG Investment Program, with AIG as the only recipient (Buchholtz and Lawson 2020; GAO 2010).

Citi leadership told the FDIC that all Citi needed was expanded access to the Fed’s liquidity facilities (FCIC Report 2011). However, the FDIC dismissed this option on the basis that (a) any “incremental liquidity” would be quickly eliminated by deposit outflows and (b) Citi did not have adequate high-quality collateral to access the Fed’s facilities^FIndeed, Citi had moved only “a limited amount” of assets for pledging at the discount window in response to an impending rating-house downgrade earlier in the year (FRBNY 2009). Further, late in the evening on Saturday, November 22, Citi requested that the Fed double Citi’s access to the Commercial Paper Funding Facility, and the Fed rejected the request (SIGTARP 2011a). (FCIC Report 2011, 381). On Saturday, November 22, 2008 (the day before the USG announced the new asset guarantee and recapitalization package for Citi), the government considered other options,^FRelatedly (though options considered by Citi instead of the USG), Citi was also considering on November 22 the sale of its whole business or parts of its business, as well as the replacement of its CEO (FactSet 2008). including: (a) the creation of a government conservatorship in the style of those created for Fannie Mae and Freddie Mac; (b) the creation of a special purpose vehicle^FA special purpose vehicle is an off-balance-sheet legal entity that can hold assets that the entity providing the capital for such a vehicle may not be able to purchase or hold (SIGTARP 2011a). controlled by the government to buy troubled assets from Citi with TARP funds; (c) the creation of a public–private investment fund to buy troubled assets from Citi; and (d) further TARP capital injections. Ultimately, the USG parties said that the ring-fencing (asset guarantee) strategy would require far less up-front funding from the USG than purchasing assets outright and would be replicable for other institutions that could come under similar pressures (SIGTARP 2011a). According to the Treasury, none of the other options were “seriously considered” (COP 2009b).

However, after the TIP investment, when market sentiment toward Citi did not sufficiently improve (resulting, ultimately, in the Treasury’s announcing a plan to exchange CPP preferred shares for common stock on February 27, 2009), nationalization was considered seriously by the Treasury and the economic leadership team at the White House^FIn fact, Obama advisers Larry Summers and Jeremy Stein thought that nationalization was both necessary and inevitable (Geithner 2014, 308). Then–Treasury Secretary Geithner and Summers wrote to President Obama saying that there was a “significant chance” that both Citi and BofA would ultimately be nationalized (Geithner 2014, 308–309). (Geithner 2014, 307–309). The Treasury team ultimately concluded that the most likely outcome of a premature nationalization of Citi would be a run on BofA, followed potentially by runs on other large banks such as Wells Fargo, Morgan Stanley, and Goldman Sachs (Geithner 2014, 325). Further, policymakers at the time agreed that if Treasury were to nationalize Citi or BofA, it would then have had to guarantee the rest of the system to prevent system-wide runs, which would have been very costly. In addition to the first-order cost of the adjacent guarantees that would have been necessary, nationalization would have presented the government with the choice of either liquidating the firm(s) quickly, resulting in expensive fire-sale losses for taxpayers, or running the institution(s) for the medium to long term, which would have introduced other challenges (Geithner 2014, 344).

Both the TIP capital injections for Citi and BofA were constituent parts of larger USG support packages, which also included the AGP (also known as the “ring-fence”), which guaranteed assets on the banks’ balance sheets in loss-sharing agreements with the Treasury, FDIC, and Fed (Arnold, forthcoming-a; Arnold, forthcoming-b).

On November 23, 2008, the Fed, FDIC, and Treasury announced an interagency support package for Citi, which, in addition to the TIP capital injection, consisted of a guarantee by the USG of $306 billion of assets (later reduced to $301 billion) (SIGTARP 2011a). Under the Asset Guarantee Program, the Treasury and FDIC agreed to guarantee ring-fence-eligible assets by taking second- and third-loss positions, respectively. Under this arrangement, Citi would bear the first $29 billion in losses (later expanded to $39.5 billion in the eventual Master Agreement)^FOriginally, the November 23 announcement stated that Citi’s deductible would be $29 billion “plus reserves”; factoring in (1) reserves with respect to the portfolio of roughly $9.5 billion and (2) an additional agreed-upon $1 billion in reserves as compensation for excluding some hedging products, the total Citi deductible at the time of signing the Master Agreement came to $39.5 billion (COP 2009b). on the asset pool, and then Treasury would provide $5 billion in loss guarantees and the FDIC would provide $10 billion in loss guarantees (Fed 2008b; Citi, Treasury, FDIC, and FRBNY 2009a). 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). For the AGP, the Treasury also received warrants to purchase of 66.5 million common shares at a strike price of $10.61 per share; for the TIP, the Treasury received warrants to purchase 188.5 million common shares at the same strike price (SIGTARP 2011a).

The Fed would provide an emergency loan for 90% of the amount of losses over $56.2 billion through a loan facility to Citi through the Federal Reserve Bank of New York. Any lending through the loan facility would be collateralized against the asset pool, which would comprise loans and securities backed by residential or commercial real estate (and other assets that the government and Citi would agree as eligible) (Fed 2008a). See Arnold (forthcoming-b) for a separate case study on the Fed’s loan facility with Citi.

For Citi, the value of the asset guarantee was that the regulators allowed the bank to report significantly higher regulatory capital ratios—which they variously reported as $16 billion of additional capital or a 150-basis-point-higher Tier 1 capital ratio. Market participants responded favorably to the Asset Guarantee Program, even though Citigroup remained responsible for the first $39.5 billion in losses, because of the widespread fears of catastrophic losses at the height of the Global Financial Crisis of 2007–2009 (GFC). On December 23, 2009, Citi and the government parties terminated the Master Agreement governing the loan facility, thereby ending the FRBNY’s loan facility (Citi, Treasury, FDIC, and FRBNY 2009b).

Earlier, in October 2008, Citi was one of 12 financial institutions to receive Treasury investments under a broad-based capital injection program through TARP, the CPP. On October 28, Citi received $25 billion, the maximum allocation under the CPP (SIGTARP 2009).

By the time of the November 23 support package, Citi had utilized several existing GFC-era programs and facilities. 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; and had borrowed $84 billion from Federal Home Loan Banks (FCIC Report 2011). On November 23, 2008—concurrent with the announcement of the USG support package—the Fed authorized the extension of credit to Citi’s London-based broker-dealer under the Primary Dealer Credit Facility (PDCF) (Fed 2009). By December, Citi had also issued $32 billion in senior debt guaranteed by the FDIC (FCIC Report 2011).

The Emergency Economic Stabilization Act (EESA) of 2008 authorized the Treasury to establish the TIP under TARP (SIGTARP 2009). President George W. Bush signed EESA on October 3, 2008, to give the Treasury novel authority to purchase distressed assets related to mortgages, securities, and other financial instruments (Baker Donelson 2008). Section 101(a)(1) of the EESA states that the Treasury was authorized to establish TARP to “purchase, and to make and fund commitments to purchase, troubled assets from any financial institution” (EESA 2008, sec. 101[a][1]). The EESA defines a “troubled asset” as (a) residential or commercial mortgages, or other financial instruments based on or related to mortgages; or (b) any instrument that the Treasury secretary, after consulting with the Fed chair, determines “the purchase of which is necessary to promote financial market stability,” after transmitting that determination to appropriate Congressional committees (EESA 2008, sec. 3[9]). Section 113(d) of the EESA stipulates that, in general, the Treasury—when acquiring troubled assets of a financial institution—must also acquire warrants for the right to receive nonvoting common stock or preferred stock of the financial institution for the benefit of taxpayers (EESA 2008, sec. 11[d][1]). Before TIP, the CPP was a broad-based program through which Treasury purchased equity in financial institutions; as a TARP facility, the CPP was also authorized under the EESA (Treasury n.d.b).

There was certain ex post controversy about the legal authority of the TIP. The Congressional Oversight Panel said:

There is no evidence that the TIP existed as a program prior to [its January 2, 2009] announcement, but funds were disbursed to Citigroup that were later attributed to the TIP . . . TIP is not referred to by name in EESA [but] Treasury asserts its authority for this program arises from Section 101, which authorizes Treasury to purchase troubled assets. (COP 2009b)

Section 101(d) of the EESA required Treasury to provide a program description of its TARP program within two business days of the first purchase of troubled assets. The EESA required that program description to include (a) the mechanism(s) through which Treasury would buy the assets; (b) the pricing/valuation methodology to be employed; (c) asset manager-selection process; and (d) criteria for selecting the troubled assets (EESA 2008, sec. 101[d]). Two days after its first capital injection into Citi, Treasury provided its program description and said that the underlying legal authority for the TIP was the EESA (SIGTARP 2011a; Treasury 2009b). In its Section 101(d) description, the Treasury said that it,

May invest in any financial instrument, including debt, equity, or warrants, that the Secretary of the Treasury determines to be a troubled asset, after consultation with the Chairman of the Board of Governors of the Federal Reserve System and notice to Congress. (Treasury 2009b)

In exercising its TARP authorities, the Treasury secretary was required by the EESA to consider, inter alia, when purchasing assets or instruments from an individual institution, that institution’s long-term viability; protecting taxpayer money and maximizing returns; and providing systemic financial stability and the stability of the broader economy, to include employment and foreclosures (EESA 2008, sec. 103).

Pursuant to Section 105(a) of the EESA, Treasury was required to make monthly disclosures about its use of TARP funds (see Key Design Decision No. 6, Communication and Disclosure) (EESA 2008, sec. 105[a]).

Citi was not involved in any discussions about the TIP capital injection; rather, the idea was suggested by the USG to Citi at the end of “Citi weekend” (SIGTARP 2011a). On Saturday, November 22, 2008, at 3:36 a.m., Citi delivered a proposal for government assistance to the FRBNY. At about noon, the USG parties (Treasury, FDIC, and Fed) met with Citi to discuss the proposal, choosing ultimately to revise the terms. That evening, Chairman Bernanke and Secretary Paulson discussed whether to make a common stock or preferred share injection. On Sunday, they submitted to Citi a new proposal with a revised term sheet (SIGTARP 2011a). Figure 3 shows the difference between the original Citi proposal (Nov. 22) and the ultimate USG support package (Nov. 23).

The main changes from Citi’s proposal were: (a) the government proposal required Citi to accept a significant first-loss position on the asset pool, whereas Citi has proposed that the government would cover all losses on the pool; (b) in return, the government would inject an additional $20 billion in preferred equity through the program later known as TIP (see Figure 3).

Source: SIGTARP 2011a.

After the July 2009 Citi exchanges (see Key Design Decision No. 8, Capital Characteristics), the Treasury said that it expected to hold the securities in a trust (Treasury 2009a).

The Government Accountability Office (GAO) provided oversight of TARP. Congress also created a Special Inspector General for TARP, which issued audit reports discussing the financial support provided to BofA as part of its broad oversight of TARP programs.

Pursuant to Section 116 of EESA, the GAO provided oversight on the performance of TARP and its related programs (including the TIP) (EESA 2008, sec. 116). The GAO subjected TARP to oversight along the following domains:

• TARP’s performance in meeting the purposes of EESA in foreclosure mitigation, cost reduction, financial stability, and taxpayer protection;
• TARP’s financial condition and internal controls;
• Characteristics of transactions and commitments;
• Characteristics and disposition of acquired assets;
• Efficiency of the use of appropriated funds;
• Legal and regulatory compliance;
• TARP’s efforts in mitigating conflicts of interest;
• Efficiency of contracting procedures. (EESA 2008, sec. 116)

Additionally, the Treasury would reimburse the GAO for oversight costs (EESA 2008, sec. 116). EESA required the GAO to submit oversight reports every 60 days to Congress and the SIGTARP (EESA 2008, sec. 116). EESA also required TARP to issue annual reports of its financial statements, audited by the GAO, to Congress and the public (EESA 2008, sec. 116).

The Inspector General (for SIGTARP) was appointed by the president and confirmed by the Senate (EESA 2008, sec. 121). SIGTARP was responsible for conducting, supervising, and coordinating audits of the purchase, management, and disposition of assets under TARP (EESA 2008, sec. 121).

On January 13, 2011, SIGTARP issued a 77-page audit report entitled “Extraordinary Financial Assistance Provided to Citigroup, Inc.,” which covered, inter alia, the basis for the decision to support Citi with the TIP and the basis for the decision to allow Citi to repay its TIP capital injection (SIGTARP 2011a).

The Treasury injected capital into Citi on December 31, 2008, before announcing the name of the official program in which the Citi intervention would be included, the TIP, on January 2, 2009 (SIGTARP 2011a). Although the capital injection would not be made until December, the USG expressed the hope that its announcement in November would send a stabilizing message to reassure Citi counterparties, which the USG said would be expected to increase lending to Citi and relax the terms thereof (SIGTARP 2011a).

On December 14, 2009, the same day Citi received approval from the Fed to repay its TIP investment and cancel its AGP arrangement, Citi issued a press release announcing as much (SIGTARP 2011a). Two days later, on December 16, Citi issued a press release announcing its planned public equity offering and saying that, with its planned repayment of TIP and its exit from the AGP, it would “no longer be deemed to be a recipient of ‘exceptional financial assistance’ under TARP” (SIGTARP 2011a, 57).

Pursuant to Section 105(a) of the EESA, Treasury was required to make monthly disclosures to Congress regarding its use of TARP funds (“Monthly 105(a) Reports”), including those used in the TIP (EESA 2008, sec. 105[a]). Starting in October 2008, Treasury has filed Monthly 105(a) Reports each month, through August 2023 at the time of writing.^FTreasury’s Monthly 105(a) Reports can be found through the Federal Reserve Bank of St. Louis: https://fraser.stlouisfed.org/title/troubled-asset-relief-program-tarp-5167?browse=2000s. Additionally, Section 129 of the EESA required the Fed to report to the Senate Banking Committee and House Financial Services Committee on any use of its Section 13(3) authority within seven days of invoking that authority (EESA 2008, sec. 129[a][2]).

During the second hearing of the Congressional Oversight Panel on Citi and TARP, panel member Damon Silvers said that the Citi TARP programs, in aggregate, resulted in a de facto restructuring of Citi’s balance sheet:

[Treasury] has managed TARP’s holdings in Citigroup to affect what is essentially a limited balance sheet restructuring; effectively requiring or inducing Citigroup’s preferred stockholders to become common stockholders—not just the government, but the private preferred stockholders that preceded the government on the balance sheet. Now, this step, this move, diluted common stockholders’ share of future profits substantially, and this is something I strongly support from the perspective of fairness and moral hazard . . . But these transactions did not appreciably alter Citigroup’s total Tier 1 capital, as a percentage of Citigroup’s risk-adjusted assets, nor did it result in any consequences for Citigroup’s bondholders. [emphasis added] (COP 2010)

Preferred Shares

Citi issued Treasury 20,000 fixed-rate cumulative perpetual preferred shares with a liquidation value of $1 million per share, for a total of $20 billion (Citi and Treasury 2008). The preferred shares were ranked pari passu with or senior to all other preferred shares in Citi, whether outstanding or not yet issued (Citi and Treasury 2008). The preferred shares qualified as Tier 1 capital (Citi 2008).

On Saturday, November 22, 2008, Chairman Bernanke and Secretary Paulson discussed whether to structure the injection as a common-share or preferred-share injection. While Bernanke initially supported a common equity investment, Paulson opted to support the preferred-share investment because it would avoid diluting common shareholder equity and the political ramifications of a large government stake in Citi. According to one senior Fed official, the use of nonconvertible preferred shares would deter market characterization of Citi as having been nationalized, which would have further undermined market confidence in the bank. Additionally, preferred shares had other advantages of (a) more seniority in the capital stack and resulting higher protection of taxpayers in the event of liquidation; (b) the payment of dividends by Citi to the USG (SIGTARP 2011a).

The preferred shares paid a cumulative 8% annual dividend, computed on a 360-day basis, payable in arrears on the 15th of February, May, August, and November. The 8% rate was calculated cumulatively on (a) the liquidation amount per share and (b) the amount of accrued and unpaid dividends of any prior dividend period. When dividend payments were not paid in full on the respective payment date, the dividends would be “declared pro rata so that the respective amounts of such dividends declared shall bear the same ratio to each other as all accrued and unpaid dividends per share” (Citi and Treasury 2008).

Per the agreement, the Treasury would not sell the preferred shares unless it were to a qualified institutional buyer. Additionally, Citi would be obligated to provide notice to Treasury of its intention to redeem preferred shares at least 30 days before—but not more than 60 days before—its planned redemption date (Citi and Treasury 2008). Citi could not redeem the TIP shares until it redeemed all $25 billion preferred shares it received under the CPP (Citi 2008). CPP participants could redeem shares after three years or upon completing a qualified equity offering—raising additional Tier 1 or Tier 2 capital through private markets equal to at least 25% of the Treasury’s investment (Lawson and Kulam 2021).

The TIP preferred shares were nonvoting but did confer class voting rights on the following: (a) the issuance of shares senior to the TIP preferred shares; (b) amendments to the rights of the TIP preferred shares; and (c) mergers, acquisitions, or other exchanges that could adversely affect the rights of the TIP preferred shareholders. Further, the Treasury would retain the right to elect two directors to the board of Citi in the event that dividends were not paid in full for a total of six dividend periods (regardless of whether those periods were consecutive), but that election right would terminate at any time at which Citi had paid all prior dividend periods (Citi and Treasury 2008).

Treasury had no right to convert the preferred shares into any other securities (Citi and Treasury 2008).

Warrants

For the TIP, Citi issued warrants to Treasury to purchase 188,501,414 shares of common stock at a strike price of $10.61.^FCiti and Treasury could reduce this price though, for example in the case that Citi completed a pro rata repurchase of common stock (Citi and Treasury 2008). For the AGP, Citi also issued warrants to Treasury to purchase 66,531,728 common shares at a strike price of $10.61 (SIGTARP 2011a). The warrants would expire after 10 years. The Treasury would not exercise voting rights with respect to shares acquired under the warrants. In the event of registration with the Securities and Exchange Commission, Treasury agreed to a 30-day holding period between notifying Citi of its intent to sell the warrants and the day on which it could sell them. The warrants were fully transferrable, in whole or in part (Citi and Treasury 2008).

Upon the redemption (in whole) of the preferred shares, Citi could repurchase the warrants from Treasury at fair market value (Citi and Treasury 2008).

Trust Preferred Securities

On February 27, 2009, the Treasury announced its intention to participate in Citi’s planned exchange offers, aimed at boosting Citi’s tangible common equity (Treasury 2009d). According to the plan, Citi would convert a significant portion of its preferred shares into common equity. Under the exchange, the Treasury would convert its CPP preferred shares—but only the amount of preferred shares that other preferred shareholders were willing to convert (up to $25 billion)—into a new security that would convert into common equity. The Treasury would also exchange its $20 billion TIP Citi preferred shares into new trust preferred securities (TruPS) with the same 8% cash dividend rate. The TruPS would have greater structural seniority (Treasury 2009d). On July 30, 2009, Treasury exchanged its preferred stock obtained under the TIP and the AGP^FThe FDIC similarly converted its $3 billion in preferred shares to $3 billion in trust securities (SIGTARP 2011a).  for new TruPS, thereby strengthening some of Citi’s capital ratios. These securities were more structurally senior to preferred stock^FIn at least one instance, Treasury noted that Citi had outstanding enhanced trust preferred securities, but research was unable to obtain further information on those securities and found no evidence that Treasury ultimately obtained any of them (Treasury 2009a). (for example, they had a more senior claim in bankruptcy), and Treasury continued to collect its 8% annual dividend. Those trust securities were set to mature in 2039 (SIGTARP 2011a).

TruPS were hybrid securities issued by bank holding companies (BHCs) through the following structure: the BHC created a special purpose entity (SPE), which issued cumulative preferred stock; the BHC then borrowed on those proceeds using long-term subordinated debt. Since a 1996 Fed decision, a portion of TruPS issuance satisfied Tier 1 capital requirements. The dividend payments on TruPS were tax deductible to the BHC (since dividend payments were in fact the interest on the subordinated debt). This was not true of dividends paid on normal preferred equity, and therefore TruPS represented a tax advantage to the issuer. While regulators considered TruPS equity for the purposes of regulatory capital, rating agencies rated TruPS as debt instruments and discounted their contributions to the capital strength of banks (Moody’s Investors Service, for example, said that it “[had] always considered TruPS to be far more debt-like in nature”) (French et al. 2010, 5). The FDIC said that in its experience, TruPS holders “have been an impediment to recapitalizations or sales of troubled banks” (French et al. 2010, 12).

The regulatory treatment of TruPS as Tier 1 capital was subject to numerous Fed conditions, including, in particular, the requirement that the eligible TruPS must confer to the BHC the ability to defer dividend payments. In the Supervisory Capital Assessment Program (SCAP) stress tests in the spring of 2009, bank regulators did not include TruPS in bottom-line tangible equity targets of the assessments (French et al. 2010).^FThe Dodd–Frank Wall Street Reform and Consumer Protection Act of 2010 eliminated most BHCs’ ability to include TruPS in Tier 1 regulatory capital levels (French et al. 2010).

According to the Treasury, the TruPS’ “material terms . . . [would] be substantially similar to those of Citi’s traditional TruPS” (Treasury 2009a). Treasury said that the broader exchange (to include the exchange of preferred shares for common stock under the CPP) did not increase Treasury’s investment in Citi but did “increase the risk of the investment for the government” (Treasury 2009d); while the trust preferred securities were more senior in the capital stack vis-à-vis the preferred shares, the common stock, exchanged through the CPP, was more junior.

The Citi TIP capital injection totaled $20 billion. Then-Secretary Paulson said his team conducted no Citi-specific analysis to arrive at the figure but rather considered the limited amount of available TARP funds and the possibility of another bank’s needing TARP funds. Although there was no written documentation concerning the process of arriving at the $20 billion figure, then–FRBNY President Geithner said the decision was the result of FRBNY and Treasury discussions (SIGTARP 2011a). In regard to the $20 billion size, Geithner said, “there’s no perfect science to this thing. . . You need to balance risk versus what the firm needs, but it was Treasury’s money” (SIGTARP 2011a, 22).

On October 28, 2008, Citi received a $25 billion capital injection, also through preferred shares, from the Treasury under the broad-based Capital Purchase Program (SIGTARP 2009). Despite the investment, Citi’s stock price continued to collapse in the following weeks, falling from $13.99 on November 3, 2008, to $3.05 on Friday, November 21 (SIGTARP 2011a). Over the weekend of November 21, Citi executives met with Federal regulators to work out a rescue plan (SIGTARP 2011a).

On Sunday, November 23, 2008, in order to support the stability of financial markets, the government announced the support package for Citi (USG 2008). The package consisted of (a) capital injections via the Treasury purchase of $20 billion in senior preferred stock through the Targeted Investment Program (Fed 2008b) and (b) a guarantee by the AGP of $306 billion of assets (SIGTARP 2011a).

After the TIP capital injection on December 31, 2008, Citi’s share price continued to decline, falling below $1 per share in March 2009. According to the OCC, markets viewed the aggregate $45 billion in TARP preferred shares as equivalent to debt and wanted Citi to obtain more common equity. By February 2009, Treasury was aware that Citi needed more capital—specifically common equity—but that rumors that the government was going to nationalize Citi were also destabilizing the financial system. In order to prevent a Citi failure on the one hand and avoid nationalization on the other hand, Treasury settled on a conversion of CPP preferred shares into common shares, but an amount that kept the government’s share of Citi below 50% (Geithner 2014, 314). On February 27, 2009, Citi and Treasury announced that Treasury had agreed to exchange $25 billion of preferred shares from the earlier CPP investment into common stock at a price of $3.25 per share and that Treasury would exchange its TIP and AGP preferred securities for trust preferred securities (SIGTARP 2011a; Treasury 2009c). Citi and the Treasury entered into the resulting agreement on July 9, 2009, and on July 23, 2009, and July 30, 2009, both parties completed the transaction, resulting in Treasury’s conversion of $25 billion of preferred shares to $25 billion of common shares (for a total of 7.7 million shares). As a result of the transaction, Treasury owned 33.6% of Citi common stock (SIGTARP 2011a).

Also on July 30, 2009, the Treasury converted its TIP preferred shares to trust preferred securities (see Key Design Decision No. 8, Capital Characteristics). Figure 4 shows the changes to Treasury’s holdings of Citi securities before and after the July 2009 exchanges.

Source: SIGTARP 2011a.

Citi did not undergo restructuring during its participation in the TIP.

Executive compensation required approval by the USG parties; see Key Design Decision No. 13, Other Conditions (SIGTARP 2011a). Under Citi’s TIP agreement, the Treasury would retain the right to elect two directors to the board of Citi in the event that dividends were not paid in full for a total of six quarterly dividend periods (regardless of whether those periods were consecutive). The election right would terminate at any time at which Citi had paid all prior dividend periods (Citi and Treasury 2008).

The preferred shares restricted common stock dividend payments and equity repurchases (Citi and Treasury 2008). For a period of three years, the Citi agreement prohibited the payment without Treasury’s consent of (a) quarterly dividends on common stock in excess of $0.01 per share; (b) any dividends payable in common stock shares; or (c) dividends or distribution rights for junior stock. Further, the Citi agreement prohibited the repurchase without Treasury consent of any equity securities (to include common stock, trust preferred securities, etc.), subject to some exceptions (Citi and Treasury 2008).

Further, executive compensation plans had to be approved by the USG (SIGTARP 2011a).

Our research did not uncover evidence of regulatory relief extended to Citi.

In compliance with the American Recovery and Reinvestment Act of 2009, Treasury had to permit repayment of any TARP capital (see Key Design Decision No. 3, Legal Authority). However, Treasury approval for repayment of TARP capital was conditional on consultation with the appropriate federal banking authority, in this case the FRBNY. Financial institutions aiming to repay TARP investments were subject to existing supervisory procedures for the approval of redemption requests for capital instruments, which included assessments by the bank regulator of, inter alia, soundness, capital adequacy, and lending capacity (SIGTARP 2011a).

On September 11, 2009, then–Citi CEO Vikram Pandit met with the FRBNY’s president at that time, William Dudley, and others about redeeming its TIP capital. The Fed was unconvinced that Citi was prepared to repay its TARP funds, particularly after running its own stress tests. On November 3, 2009, the Fed provided further guidance on steps necessary for institutions to repay TARP investments, among them a demonstrated ability to access long-term debt markets in the absence of the Temporary Liquidity Guarantee Program (SIGTARP 2011a).

Two days later, an FRBNY official met with Citi’s CFO and CEO and informed them that Citi would have to increase its common share equity more than other banks had before they would allow Citi to repay its TARP capital (SIGTARP 2011a). On November 9, the FRBNY ran an additional stress test on Citi and on December 3 presented the results of that stress test to Citi’s chief risk officer and CFO. To be eligible for TIP repayment, Citi would need to maintain a Tier 1 capital ratio of 6.6% and a Tier 1 common ratio of 4.6%, which would be possible if Citi: (a) raised $21.4 billion in common stock; (b) repaid the $20 billion in TIP trust preferred securities; and (3) canceled the AGP ring-fence arrangement. Later in the day (on December 3), Citi provided the FRBNY with a repayment proposal, which included full repayment of the TIP capital injection and termination of the AGP. In its proposal, Citi said it would raise $20.3 billion of capital. The FRBNY responded to the proposal saying that it did not include sufficient common equity issuance. The following day, Citi revised its proposal, raising the total amount to $24.1 billion (SIGTARP 2011a). Figure 5 shows the composition of the original and revised repayment proposals.

In response to Citi’s December 10 proposal, the FRBNY said that it was concerned about Citi’s ability to fulfill the overallotment option and said that Citi’s final proposal would need to include a stipulation about actions Citi would take in the event the overallotment was not filled (SIGTARP 2011a).

On December 13, 2009, Citi submitted a final repayment proposal to the FRBNY, which matched the December 10 proposal in capital instrument classes and sizes but included the stipulation that the USG would surrender $1.8 billion of AGP trust preferred securities as compensation for the early AGP termination^FThis figure was the result of independent negotiations with the Treasury (SIGTARP 2011a). and acknowledged that it would need to make up any capital shortfall in the event it was unable to raise sufficient capital:

. . . if the offering of common stock and tangible equity units do not generate at least $21.3 billion of additional equity capital, the regulators would expect Citigroup to issue additional trust preferred securities in a ratio of $2 for every $1 the equity raised falls short of $21.3 billion, subject to a minimum equity raise of $19.8 billion, up to a maximum of $3 billion of trust preferred securities during the first quarter of 2010. (SIGTARP 2011a, 37)

On December 14, the Fed sent a letter to Citi approving its final repayment proposal and outlining the conditions necessary for Citi to exit the TIP and the AGP. Two days later, Citi priced its capital offering and began to execute the capital raise. It was the largest public equity offering in US capital market history at the time (SIGTARP 2011a).

On December 23, 2009, Citi and Treasury signed an agreement for the repayment of the TIP. While the Termination Agreement said that Citi had raised capital through common and preferred stock offerings, in fact, Citi had failed to meet the $1.5 billion overallotment option necessary for fulfilling the capital requirement, raising only $0.6 billion and thereby leaving a $0.9 billion capital shortfall. In March 2010, Citi raised $2.3 billion through an offering of trust preferred securities (SIGTARP 2011a).

Throughout 2010, Treasury began to sell its AGP trust preferred securities and CPP common shares—by December 10, 2010, Treasury had sold all its CPP common stock and the AGP and TIP trust preferred securities. However, as of December 10, 2010, Treasury maintained warrants for 465.1 million common shares and had finished liquidating CPP common shares (SIGTARP 2011b; SIGTARP 2011a). By January 25, 2011, Treasury sold its outstanding warrants from the Citi TIP for a total proceed of $190.4 million (Treasury 2015).

Between the disposition of warrants and dividend payments, the Treasury received $1.8 billion (excluding interest expense) in income from the Citi TIP investment (Treasury 2013; Treasury 2015). Combined with the investments into BofA, the TIP generated a profit of $4 billion in total, with $3 billion in dividend payments (from $40 billion in investments) (Treasury n.d.a).