TARP, Emergency Intervention, and Equity Stakes

Paulson later described the instinct behind that moment with unusual candor: “Even if you don’t have the authorities—and frankly I didn’t have the authorities for anything—if you take charge, people will follow. Someone has to pull it all together.” That comment captures both the practical and the separation of powers problems. In a panic, officials may act first, find authority later, and leave taxpayers, markets, and Congress to sort out the consequences.

This essay traces the causes, justifications, and operational details of the Troubled Asset Relief Program (TARP) and related emergency support programs. The central question is not whether the 2008 crisis was serious—it was. The question is whether those emergency interventions provide a sound precedent for normalizing government ownership, capital injections, and federal dealmaking in private firms today. They do not.

A severe recession began in December 2007, though the depth of the financial instability was not clear in real time. The turbulence reached a crescendo in the second half of 2008, concentrated in housing finance and mortgage markets after years of bipartisan policy that encouraged borrowing and lending. The politically favored metric was the US homeownership rate. That policy coincided with accommodative monetary policy in the early 2000s and a breakdown in oversight by the alphabet soup of federal financial supervisors, including the Federal Reserve, Federal Deposit Insurance Corporation (FDIC), Office of the Comptroller of the Currency (OCC), Office of Federal Housing Enterprise Oversight, and Office of Thrift Supervision.

By 2008, the crisis had moved from warning signs to emergency responses. The Federal Reserve Bank of New York helped arrange the ad hoc bailout of Bear Stearns in March, and the Federal Reserve Board announced emergency support programs. IndyMac Federal Savings Bank failed in July at an estimated cost of $12 billion, then the highest-cost bank failure in FDIC history. In September, Fannie Mae and Freddie Mac entered government conservatorship; Lehman Brothers and Washington Mutual failed; and American International Group received extraordinary support. Shortly thereafter, Congress enacted TARP, and the Federal Reserve and FDIC developed additional emergency support programs.

Consistent with the idea that “there is nothing new under the sun,” several precedents had laid the foundation for emergency financial support during wars and financial crises. The War Finance Corporation Act of 1918 provided funding to enable banks to furnish credit for industries and enterprises necessary to prosecute World War I. The war ended that year, but the War Finance Corporation (WFC) soon had its mandate expanded to support agricultural banks after farm prices fell.

President Herbert Hoover seized upon the WFC wartime precedent, developing a peacetime government bank-lending program during the Great Depression: the Reconstruction Finance Corporation (RFC). When RFC lending powers failed to revive the depressed economy, President Franklin D. Roosevelt and Congress expanded the RFC’s responsibilities to provide capital contributions in the form of preferred stock under the Emergency Banking Relief Act. That allowed the government to become an incredibly influential shareholder.

Continental Illinois of Chicago is perhaps the starkest illustration. The RFC’s $50 million equity stake in 1933 gave it control of the bank, and the RFC leveraged that control in 1934 to appoint Walter J. Cummings, Roosevelt’s appointment to the FDIC board, as chairman of Continental’s board. The RFC also used its shareholder power to restrict executive compensation and dividends at the banks receiving funds.

The pattern has become familiar: An emergency finance tool becomes an ownership tool, and ownership becomes a governance tool. Continental Illinois reprised its role as bailout recipient in 1984 when the institution stumbled again, and the FDIC acquired an 80 percent ownership stake and installed retired Standard Oil executive John Swearingen as chairman of the holding company.

As the financial crisis deepened in the fall of 2008, the federal government’s response was increasingly ad hoc, late-night, and seat-of-the-pants. The most troubling trend was the encouragement of shotgun marriages (i.e., mergers) among financial institutions. After Bear Stearns, Paulson told Bank of America chief executive Ken Lewis “to seriously consider buying [Lehman]” and to put on his “imagination hat” to come up with a deal. Paulson and New York Federal Reserve Bank President Timothy Geithner suggested at least 10 such mergers, creating uncertainty across the industry. Bankers, in response, derisively called Geithner “eHarmony,” after the dating service.

The Treasury then turned to a broader response backed by new legislation. Assistant Secretaries Neel Kashkari and Phillip Swagel dusted off the April 2008 “Break the Glass” Bank Recapitalization Plan. The primary approach directed the government to buy illiquid mortgage-related assets from banks so they could resume lending and stabilize housing and mortgage markets. Its major components included purchasing $500 billion of mortgage-backed securities from US-headquartered financial institutions, compensating bidders with Treasury securities rather than cash, hiring private asset managers to maximize value for taxpayers, unwinding positions over time, and allowing the government to receive 10 percent equity in the institution.

Four alternatives were presented, one of which involved Treasury funding new equity in banks. That option was initially treated as the least plausible alternative. During the legislative push, the asset purchase figure rose to $700 billion. According to one Treasury spokeswoman, “We just wanted to choose a really large number.” Fear tactics also became part of the sales pitch. Kashkari said of congressional staff, “We’ve got to scare the shit out of the staff.” Sen. Jon Tester (D‑MT) pushed back at a Senate hearing: “Why do we have one week to determine $700 billion that has to be appropriated or this country’s financial system goes down the pipes?” President Bush echoed Treasury’s statements in a national address, warning that “without immediate action by Congress, America could slip into a financial panic.”

Even before TARP was finalized, it was becoming clear that the troubled-asset purchase plan was unworkable. Paulson and Kashkari had overpromised on what they could deliver. They did not fully comprehend how difficult it would be to select and purchase assets from operating institutions. After the Emergency Economic Stabilization Act of 2008 passed, the Treasury completely changed directions. Some observers called it a “bait and switch.”

Instead of buying troubled mortgage assets, Treasury chose RFC-style capital purchases to push banks to loosen standards and jump-start lending. That was not the core plan sold to Congress. Nor was bank capital the obvious binding constraint. The Federal Reserve’s Senior Loan Officer Opinion Survey from the week of TARP’s signing showed that deterioration in bank capital was not one of the top reasons for banks tightening credit standards and loan terms. As Table 1 shows, it was actually the uncertain economic outlook.

The tools granted under TARP included purchases of troubled assets, guarantees, loans, and warrants. The purchase of preferred stock was not explicitly mentioned as a stand-alone power, but the Treasury absorbed it within the statute’s broad definition of troubled assets as “any other financial instrument” the secretary deemed necessary to promote financial market stability. Table 2 shows a breakdown of the $475 billion in Treasury commitments across the headline program categories, as the $700 billion cap was reduced under the subsequent Dodd–Frank Act.

Most of the roughly $250 billion committed to banking institutions consisted of preferred-stock purchases: $205 billion through the Capital Purchase Program across 707 institutions, $40 billion ($20 billion each) through the Targeted Investment Program for Citigroup and Bank of America, and $5 billion for Citigroup through the Asset Guarantee Program. The Congressional Budget Office estimates that the Capital Purchase Program produced a $16 billion net gain and that the targeted investments and related guarantees produced another $8 billion. Treasury also reported gross proceeds of more than $9 billion from warrant dispositions.

Through TARP and related interventions, the federal government also acquired large common-equity positions in five major companies, primarily by converting loans or preferred shares. Table 3 shows those ownership stakes as of June 1, 2010.

The auto industry is a reminder of how quickly emergency authority can drift and lead to mission creep. TARP was sold principally as a financial stability program for the financial industry. No mention of the automobile industry appeared in the statute, a point Sen. Richard Shelby (R‑AL) raised at a TARP hearing. By contrast, the Carter administration’s bailout of Chrysler in 1979 rested on explicit legislation.

Media accounts often treated TARP as the sole source of bailouts and implied that a few hundred billion dollars in costs was a bargain. That framing badly understated the broader emergency architecture. In 2019, the Washington Post fact-checked Sen. Bernie Sanders’ (I‑VT) claim that Wall Street “got a trillion dollar bailout.” The Post counted only TARP’s $45 billion for Citigroup, ridiculed Sanders’ math, and awarded him two Pinocchios. That breadth of analysis was far too narrow.

A better estimate of Citigroup’s bailout tab is $517 billion, not $45 billion (Table 4). The data came from an FDIC internal report, as of January 31, 2009, on Citigroup’s liquidity and funding position. If Citibank, the underlying subsidiary bank, had been allowed to fail and go through FDIC receivership, then the FDIC would have had to clean up the mess, so its internal accounting is a useful source. Treasury also received an offsetting $1.8 billion, excluding interest expenses, from warrants and dividend payments flowing from the bailout.

Many Citigroup programs were announced on November 23, 2008. The government bent over backward to accommodate the firm’s funding needs. The same FDIC report cited a Citigroup analysis that the firm would be insolvent by November 26, 2008. Then–FDIC Chair Sheila Bair expressed interest in taking Citi into receivership, an option also mentioned in Sen. Elizabeth Warren’s (D‑MA) memoir. Receivership could have eliminated or reduced the government’s risk and avoided making the government a preferred shareholder.

Government officials had substantial discretion. The Capital Purchase Program, a TARP initiative, used standardized investment terms and a multilayered review process. But the Emergency Economic Stabilization Act, which created TARP, relied on broad considerations such as financial stability, taxpayer protection, and a recipient’s long-term viability. Those considerations were difficult to assess simultaneously during a crisis, and the statute left the Treasury secretary to define eligible instruments and set transaction terms.

The Treasury and bank regulators made case-by-case viability judgments using confidential information from on- and off-site examinations and internal recommendations. For banks with less than $1 billion in assets, the standard was whether they were well or adequately capitalized as of June 30, 2008. Very few institutions were turned down. Colonial BancGroup, for example, applied for $570 million in TARP funds but never received the funding after employees were found to have attempted to defraud the program.

On October 14, 2008, Treasury Secretary Paulson announced that the first nine large banks would receive the initial allocation of TARP money: “Government owning a stake in any private U.S. company is objectionable to most Americans—me included.” He added that “these are healthy institutions” taking this step for the good of the economy. Citigroup should have undergone a rigorous viability test consistent with TARP legislation. Less than six weeks later, the financial authorities were putting together a plan to prevent Citigroup’s meltdown.

The supervision of Citigroup failed badly. Formal communications from the OCC and Federal Reserve Bank of New York in early 2008 noted operational deficiencies at Citigroup and its core bank, while an informal memorandum of understanding enumerated other concerns. Bair summarized the breakdown bluntly: “Virtually no meaningful supervisory measures had been taken against the bank by either the OCC or the NY Fed.”

Smaller weak banks also received TARP funds. Pacific Coast National Bank is a useful example. Its holding company applied for over $4 million in TARP support in October 2008, and the Treasury approved the disbursement based on a viability test. The money was disbursed in January 2009. A senior OCC official later said that shortly after the disbursement, “the wheels fell off.” The Treasury Office of Inspector General report stated the obvious: It would have been prudent to reaffirm the bank’s condition before Treasury disbursed the money. The over $4 million was lost when the bank failed.

Wide discretion was matched with limited disclosure. The decision process for approving TARP support was shielded by deliberative process and confidential supervisory information doctrines. Those doctrines may have narrow, time-limited, and legitimate uses, but in this setting, they kept taxpayers from seeing how and why their money was being committed.

The Office of the Special Inspector General for TARP criticized the Treasury’s weak transparency posture in a 2009 report. It also warned that taxpayers were not being told what most TARP recipients were doing with the money, still had not been told how much their substantial investments were worth, and would not be told the full details of how their money was being invested. That was backward. The default should have been more disclosure, not less.

The public eventually saw TARP data on the “how” of spending through ProPublica’s Bailout Tracker, though it did not capture broader non-TARP support. The Federal Reserve and FDIC cloaked many of those programs in secrecy, and it took lawsuits to unearth the details. The highest-profile case was Bloomberg v. Board of Governors of the Federal Reserve, which forced disclosure of Federal Reserve lending data after a nearly three-year fight. The FDIC did not voluntarily reveal the Citigroup support documents without a years-long fight; they were released only after a 2015 Freedom of Information Act (FOIA) lawsuit.

TARP was not merely about repayment. Secretary Paulson imposed a mandate on the nine large banks: All had to participate whether they wanted the money or not. This was done to avoid attaching a stigma to recipients, but it also gave the government a foothold across the major banks.

According to former Wells Fargo chief executive Richard Kovacevich, banks that refused TARP support were threatened that their regulator would find them capital deficient. He argued that TARP caused the crisis to become “much greater,” noting that shortly after TARP, the stock market fell by 40 percent and banking stocks fell by 80 percent. Former St. Louis Federal Reserve President William Poole called the Treasury’s methods “arm-twisting” and “credit allocation.” Restrictions on dividends, share repurchases, and executive compensation followed.

This is where ownership, even a minority share, creates a distinctive problem. Government officials acting as overseers become tempted to put a thumb on the scale for institutions in which the government retains an ownership interest or governance role, demonstrating that the program worked. That can mean timely forbearance, open-ended additional capital, or other funding for favored firms. It is not a level playing field. The weaknesses of state-owned banks around the world are instructive regarding incentives, misallocation of funds, and profitability.

The need for emergency liquidity in 2008 was real. Many banks and financial institutions faced constrained funding positions as nonperforming assets rose. The question is what the government should demand in return for support. One option is a passive repayment structure. Another is the more aggressive model of warrants, preferred stock dividends, and active powers to intervene in governance. TARP epitomized the latter.

That ownership model enabled many of the problems described above. Preferred stock and warrants gave the Treasury a financial interest in recipients’ survival. That interest conflicted with the federal government’s role as an independent overseer. It may have distorted officials’ assessment of Citigroup in October 2008. It also gave officials an incentive to downplay prior supervisory failures and keep Citigroup afloat through repeated injections from multiple federal sources.

The same conflict surfaced when regulators threatened otherwise sound banks, including Wells Fargo, if they refused TARP funds. A genuinely passive, secured lending structure would not have eliminated discretion over eligibility, collateral, or systemic risk. But it would have avoided making the government a shareholder and reduced the conflicts of interest and opportunities for political control that came with ownership.

The traditional “lender of last resort” doctrine provides a better model for when liquidity support is needed. It is imperfect and still relies on central bank judgment, but it is typically limited to secured repayment at a penalty rate with over-collateralization. The central bank can protect itself against losses and price lending based on risk without taking an ownership stake or creating the same conflict of interest.

The arguments used to justify ownership and bailouts were often vague. Wachovia (acquired by Wells Fargo in 2008) provides an example. An FOIA suit was needed to obtain documents about FDIC deliberations on an approved bailout before the Wells Fargo acquisition. Internal analysis relied heavily on words such as “might,” “could,” and “likely” while offering little concrete analysis of the probability or magnitude of the feared fallout. Officials speculated that investors could doubt other institutions, liquidity pressures could increase, money market funds might break the buck, and short-term funding markets might virtually cease. That is significant, but it is not the same as a rigorous justification for government ownership or open-bank assistance.

Expectations of support have grown over time. Each new wave of intervention has ratcheted up the breadth of direct state intervention and raised expectations for future support. The wartime WFC authority led to a broader bank support model. The Great Depression–era RFC expanded the model to include capital investments and governance mandates. The FDIC, created in the 1930s, received permanent authority to provide assistance through amendments to its enabling act in 1950. The FDIC’s 80 percent ownership takeover of Continental Illinois in 1984 further embedded the bailout culture.

TARP and related programs doubled down on those experiments by backstopping a broader array of institutions. As Alex Pollock and Howard Adler put it, crisis episodes now occur “on average about once a decade.” The Federal Reserve Bank of Richmond’s Bailout Barometer estimated that 60 percent of financial industry liabilities were government-protected as of 2016, up from 40 percent in 1999.

When signing the Dodd–Frank Act in 2010, President Barack Obama promised that the law would “put a stop to taxpayer bailouts once and for all.” The bank bailouts of 2023 made that promise look hollow. The problem is not only cost. It is the expectation that Washington will socialize losses when enough pressure builds.

The common defense of TARP is that the money was nearly repaid, that taxpayers made gains on some investments, and that the financial system avoided another Great Depression. Such a verdict is much too generous. It should not be surprising that supported institutions bounced back when the government guaranteed them against every conceivable negative consequence of their actions. Any industry would look better after receiving equivalent open-ended support.

Decisionmakers are not infallible, and they do not have special knowledge about the broader economy or individual institutions. They almost never see a crisis coming. President Hoover spoke optimistically of a “new era” right at the beginning of the Great Depression. Government officials in 2008 were no exception. During his “Great Moderation” speech in 2004, Federal Reserve Chairman Ben Bernanke said recent history made him optimistic for the future, just a few years before the global financial crisis.

In his memoir, Geithner gave an honest assessment of his qualifications to be president of New York’s Federal Reserve Bank: “I was only forty-two, and I had never worked in the finance industry.… I had never supervised or regulated banks or markets.” Former Treasury Secretary Robert Rubin, a Citigroup board member, was his patron and advocate for the job. Geithner admitted that Rubin “probably gave Citi an undeserved aura of competence in my mind.” Citi then crashed and needed unprecedented support on Geithner’s watch.

For his part, Paulson described subprime troubles as “largely contained,” called the economy “very healthy” and “robust” just months before the recession, said after Bear Stearns that “the worst is likely behind us,” and after IndyMac claimed, “our regulators are on top of it.” They were not. Citigroup was even allowed to bid for Wachovia deep into 2008, just weeks before Citi’s own failure. Paulson also assured Congress that Fannie Mae and Freddie Mac were “very viable” and that he had no intention of nationalizing them just months before they were nationalized.

The policy debate about efficacy is always asymmetric. Agencies warn that disaster will follow without intervention, while outsiders have to wait years, often leveraging the FOIA, to see the underlying decision documents. The opacity continued with the bank bailouts of 2023. Meanwhile, officials write memoirs and institutional biographies arguing that their actions worked. Geithner’s biography on his Yale University faculty page says he was a principal architect of a strategy that “avert[ed] economic collapse.” The Treasury says that TARP helped prevent another Great Depression. That is a low bar. Any administration since the 1930s can claim success on that basis. The arguments that TARP was a success are not convincing.

Falsely declaring TARP a success emboldens current and future administrations to cite past intervention as precedent for similar government investment programs. That is the wrong lesson: From start to finish the 2008–2009 crisis responses from the Treasury, Federal Reserve, OCC, and FDIC were not a success. The government’s method for addressing the instability of a modest-sized bank through an exit strategy has been successful, as it is primarily a market-driven FDIC process reliant upon open market bidding for failing institutions. The process also has some good transparency aspects, as a short time after a bank is shuttered, the terms of the transaction and a matrix with details of all the bids for the bank are released.

Alternatively, when a large bank is involved in an ad hoc merger or a TARP bailout (about two-thirds of TARP funds to banks went to the large banks), any investment decisions for the government pass into the political realm. Market-based methodologies are jettisoned, and the decisions are made by politicians or their appointees with the details kept secret for many years. Political decisions about the amount of funds provided and recipients of investments cannot be trusted because they do not have to pass a market test. Nor did TARP repayment eliminate the deeper problems of discretion, favoritism, opacity, and conflicts of interest. Similarly, today’s government stakes in companies are being assembled ad hoc, one company and sector at a time, through a patchwork of questionable statutory authorities by political appointees.

Treasury Secretary Scott Bessent has helped advance the idea of a US sovereign wealth fund. Such funds have often played a role globally in direct state intervention by making investments on behalf of governments. The Kuwait Investment Authority, for example, took a 16 percent stake in Gulf Bank in 2009 after the bank struggled when a foreign-exchange bet went bad.

President Trump, like Paulson and Geithner before him, has great confidence in his deal-making skills. After his administration orchestrated a government stake in Intel, Trump boasted that he would make “deals like that for our Country all day long.” And the administration has kept going, having now announced stakes in over 30 companies. Banks may well become targets, which is precisely why the TARP precedent should be treated with skepticism rather than celebrated.

The last thing the United States needs is a sovereign wealth fund (formal or informal) or any other standing method for implementing direct state intervention, ready to inject capital into poorly managed firms the next time a crisis arrives. Emergency programs have a way of becoming precedents; precedents have a way of becoming expectations; and expectations of rescue make the next intervention easier to justify before anyone has to prove that it worked.