Too Big to Fail

Compared to the banking and savings-and-loan (S&L) crisis of the 1980s and early 1990s, the 2008 financial crisis was much deeper and broader, encompassing a much wider range of financial institutions and investment vehicles. The FDIC played an important role in limiting the damage, drawing on past experience and institutional memory. However, unlike the earlier crisis, the FDIC was not in the forefront. The Treasury and the Federal Reserve took the lead in handling most of the larger individual failures.

Authority was divided. The FDIC had responsibility for resolving failing commercial banks and S&Ls, those institutions that had federal deposit insurance. The Treasury and the Federal Reserve assumed responsibility for handling other types of failing financial institutions such as investment banks, money market mutual funds, and insurance companies, those financial institutions that did not have federal deposit insurance.

Capabilities differed as well. The FDIC had years of experience with carefully constructed laws and procedures to resolve failing institutions. The Treasury and the Federal Reserve had no such history. They were forced to improvise, relying on bailouts or the bankruptcy process, neither being a good course of action for handling large failing financial institutions.

Much has been written about the 2008 financial crisis, including the many disagreements among the key decision makers. One area of divergent opinions was about when bailouts were necessary. President of the Federal Reserve Bank of New York, Tim Geithner, and Secretary of the Treasury, Hank Paulson, believed Washington Mutual (WAMU) should have been bailed out. Sheila Bair not only believed WAMU should have been allowed to fail—she wanted to explore letting Wachovia and Citibank fail. My opinion is different. I believe we ended up in the right place. Banks like IndyMac and WAMU should have been closed. However, there were a handful of U.S. banks, Wachovia and Citigroup included, that were too interconnected with other financial institutions throughout the world. Trying to unwind their operations, particularly under the market conditions at that time, would have made the crisis much worse. They were too big to fail.

The different views expressed by the key decision makers highlight the importance of retaining checks and balances in our financial regulatory system. The issues were complex and the stakes high. No individual or organization had all the answers, and no individual or organization should have all the power. There is good reason why the federal government has a well-thought-out and carefully designed system of checks and balances. That system with three branches of government has served us well. While differences of opinion during difficult times create heated debate and much frustration, during the 2008 financial crisis, those debates and the checks on individual action produced better results.

The year 2008 began with five vulnerable large U.S. investment banks: Goldman Sachs, Morgan Stanley, Merrill Lynch, Lehman Brothers, and Bear Stearns. All of these institutions were highly leveraged and thus highly susceptible to any financial shock. Only about 3 percent of their money was their owners' capital; the rest was borrowed money. To compound their risk exposure, most of their borrowed money was in the form of short-term, overnight funding, requiring a constant return to the market to renew their borrowings.

Bear Stearns was the first to run into trouble. The firms that lent overnight money to Bear Stearns held collateral to protect themselves against the possibility of losses. But as housing prices deteriorated, they demanded additional collateral. This placed a growing strain on Bear Stearns, and its cash position rapidly weakened. Such activity can feed on itself at an ever-accelerating pace. The weaker the housing market and Bear Stearns grew, the more lenders demanded additional collateral or stopped lending to them altogether. In March, Bear Stearns's stock price plunged from over $80 a share to just over $2 in the course of a week.

The Federal Reserve and the Treasury Department determined that the bankruptcy process would be too disruptive for handling Bear Stearns. Instead, they arranged for its sale to JPMorgan Chase (JPMC) for $2 a share. Later, the price was raised to $10 a share. The Federal Reserve Bank of New York provided $30 billion in financial assistance to support the transaction, money that was backed by some of Bear Stearns's poor-quality mortgage-backed securities. While Bear Stearns's shareholders paid dearly for the firm's collapse, the firm's bondholders and other creditors were protected against any losses. Instead, it was the Federal Reserve Bank of New York that was exposed to possible losses from the transaction.

IndyMac was the next relatively large failure, but since it was an S&L, it fell under the FDIC's purview. Unlike Bear Stearns, there was no bailout. It was a much smaller firm and had far fewer interdependencies with other financial institutions, so a bailout was never considered. Without a ready buyer, the bank was shut down in mid-July and reopened under FDIC ownership. Shareholders were wiped out and uninsured creditors were exposed to losses. This created short-term turmoil, which died down when most of IndyMac's depositors realized they were fully insured, and when other bank customers across the country familiarized themselves with how to ensure that their deposits were fully insured by the FDIC.

A couple of months later, on Sunday morning, September 8, I walked to the grocery store two blocks from my apartment in Old Town Pasadena to buy something for breakfast. Putting some orange juice, English muffins, and a newspaper on the checkout counter, I noticed a worried look on the cashier's face. She looked at me and said she did not want to live in a socialist country. I thought her comment was in reference to the FDIC's temporary ownership of IndyMac (which I was still overseeing), so I tried to reassure her. She pointed to the newspaper in front of me. In big, bold letters the headline announced that the federal government had taken over Fannie Mae and Freddie Mac. These two government-sponsored enterprises, with $5 trillion in outstanding debt and guarantees, were the poster children for too big to fail. Defaulting on that much debt would have wreaked havoc throughout the world's financial markets. These behemoths technically fell under the authority of the Federal Housing Finance Agency, but the Treasury was actively involved in determining how to handle the situation. When they were taken over, with a $200 billion commitment from the federal government, their creditors were bailed out.

Lehman Brothers was in serious trouble shortly thereafter. While it wasn't nearly as large as Fannie and Freddie, by most standards it was enormous, with over $600 billion in total assets. It, too, was heavily exposed to losses from subprime mortgages and mortgage-backed securities. Its cash was nearly gone, as investors had stopped lending it money. The company's stock price dropped by 77 percent in a week. On Friday, September 12, the Federal Reserve Bank of New York held an emergency meeting with senior representatives from a number of large financial institutions to discuss Lehman's fate, but a private-sector solution failed to materialize. After all of the criticism over the government's bailout of bondholders at Fannie Mae and Freddie Mac a few days earlier, the Treasury Department seemed determined not to bailout Lehman's creditors. The 158-year-old firm declared bankruptcy on September 15, 2008.

For the first time, the bondholders at a large insolvent financial institution were not bailed out. However, unlike the FDIC's process for handling failed banks, the bankruptcy process has no mechanism for leaving contracts in place. As a result, creditors closed out their contracts and rushed to liquidate the mortgage-backed securities they held as collateral. The fire sale forced asset prices down dramatically, costing Lehman's creditors dearly. The new and lower market prices for mortgage-backed securities meant that similar assets held by other firms had to be marked down, further weakening those companies.

Lehman's bankruptcy had a cascading effect on other institutions. With asset prices in a seeming free fall, the market was no longer sure which firms still were solvent and which were not. There was a flight to safety. No one wanted to lend money, choosing to preserve cash instead. The same day that Lehman declared bankruptcy, a severely weakened Merrill Lynch sold itself to Bank of America. Of the large, independent, investment banks, Goldman Sachs and Morgan Stanley were the only two that remained. Fearing that they would be next to suffer a drain on their liquidity, on September 21, the Federal Reserve granted both institutions approval to become bank holding companies. This protected them from a cash shortfall by giving them access to the Federal Reserve's emergency lending facilities.

The regulator of these five large investment banks, the Securities and Exchange Commission, had never developed plans for how potential insolvencies should be handled, other than through bankruptcy proceedings. When problems arose, the Federal Reserve and the Treasury Department had to improvise.

AIG, a large insurance company, was next. It had sold $500 billion worth of insurance protection against losses to financial firms around the world. These insurance policies or “credit default swaps” protected their buyers against at least some of their losses if their underlying loans went bad. AIG had sold insurance protection to many of the owners of subprime mortgages, so when the market collapsed, AIG found itself on the wrong side of these subprime bets. It hadn't charged much for its insurance protection since it hadn't viewed its risk as being very high. And under the woefully inadequate Basel II capital rules, AIG didn't have to hold much capital against these types of contingent risks.

On September 15, the day Lehman Brothers declared bankruptcy, AIG's share price dropped 61 percent. Given the enormous breadth of AIG's operations and the damage caused by Lehman's bankruptcy, the Treasury Department and the Federal Reserve felt they had little choice but to bail the company out. The next day the Federal Reserve Bank of New York lent AIG $85 billion, an amount larger than the entire taxpayer cost from the RTC rescue effort. In return, the federal government obtained over 80 percent ownership of the company. The federal government's exposure to AIG soon increased to $182 billion.

Amid all of the turmoil unleashed by these events, the largest S&L in the country, with $309 billion in total assets, also was teetering on the verge of collapse. The FDIC, which up to this point had few direct responsibilities connected to the unfolding financial crisis, would take the leading role in the process of closing the bank and devising a strategy that would result in finding a suitable buyer.

The bank, Seattle-based Washington Mutual, had been a major lender in the subprime mortgage market. Much of its lending was in California, one of the states hardest hit by the collapse of the housing market. Like IndyMac, WAMU had specialized in some of the riskiest types of mortgage loans, in effect gambling that the housing market would not collapse. WAMU had originated and owned a large volume of option adjustable-rate mortgages or “option ARMs.”

Option ARMs had two features that made them popular. First, the interest rate was kept artificially low for the first two or three years. Second, rather than having a fixed monthly payment, borrowers could decide how much they wanted to pay each month. If a borrower's monthly payment was not sufficient to cover the interest owed in any given month, the difference was added on to the principal balance. These two features enabled borrowers to keep their monthly payments artificially low for a couple of years, which encouraged them to purchase more expensive homes than what they could really afford.

Their bet and WAMU's bet was that housing prices would continue to rise. This would enable borrowers to refinance their mortgage before their monthly payments increased from their initial teaser rates. Instead, when housing prices dropped, these borrowers found that they could not refinance their mortgages, and many could not afford their newly reset monthly payments. Therefore, they owed the bank much more than what their house was worth.

By early 2008, the FDIC was closely monitoring WAMU's deteriorating condition. Chris Spoth, the senior deputy for bank supervision, led the effort, working closely with Stan Ivie, who directed bank supervision in the San Francisco region. WAMU's losses were high enough that it needed to raise more capital. JPMC was interested in purchasing WAMU. They viewed its large West Coast presence and retail-banking operations as a perfect complement to their primarily East Coast and capital-market-oriented activities. But WAMU's management was not ready to give up its independence or lose their jobs. They wanted to find investors who would contribute additional capital without taking such extreme measures.

Sheila Bair was concerned about WAMU's unwillingness to engage in conversations with JPMC. She told John Reich, the director at the Office of Thrift Supervision (OTS), that given WAMU's condition, it should not be turning down discussions with any potential source of new capital. But WAMU found the type of investors it was looking for. In April 2008, the bank announced that it had raised $7 billion in new capital. A private equity firm, TPG Capital, contributed $1 billion, and other investors, including many large institutional investors who already owned some of the bank's stock, contributed the rest. For the moment, WAMU's crisis seemed to have passed. Nevertheless, given its size and the continued deterioration in the housing market, WAMU remained a serious concern at the FDIC.

In late July 2008, WAMU's CEO, Kerry Killinger, and some of the bank's senior managers, traveled from their Seattle headquarters to Washington, D.C., to meet with their regulator, the OTS. They wanted to explain why they believed the bank was going to emerge from the turmoil in stable condition. Reich invited Bair, who attended the meeting along with two senior members of her staff, Sandra Thompson and John Corston. Reich and his senior staff seemed sympathetic to the bank's position. However, Bair said that WAMU remained a big concern to the FDIC and urged Killinger to remain open to the possibility of a merger.

Reich was angry. The OTS was WAMU's primary federal regulator, and he did not feel it was appropriate for Bair to voice her concerns in front of WAMU's management without having coordinated any discussion with him. Once again, disagreements were surfacing between the FDIC and a troubled bank's primary regulator. Behind the scenes, the two agencies had been arguing about the appropriate rating for the bank. Now the dispute was being showcased in front of WAMU's management. A heated exchange took place between the two agency heads, and Reich abruptly ended the meeting.

In September, WAMU announced second-quarter losses of $19 billion. Killinger was forced to resign by his board of directors. Alan Fishman, a former president and chief operating officer at Sovereign Bank, another large thrift, replaced him. Many of the bank's depositors began to worry about the safety of their money. They withdrew $16.7 billion over a 10-day period, beginning on September 11.

This was perhaps the largest individual bank run in history, but it largely went unrecognized outside of the bank and the bank regulatory agencies. It was painfully clear to us that WAMU was running out of cash and might have to be closed on short notice. We knew that it represented a potential crisis, as WAMU was nearly 10 times larger than any bank the FDIC had ever closed.

As of June 30, 2008, WAMU had 2,300 branches and $188 billion in deposits, an estimated $45 billion of which was uninsured. By comparison, IndyMac had only about $500 million in uninsured deposits when it was closed. We were fearful that refusing to protect WAMU's uninsured depositors would be devastating to an already fragile public confidence. But, by law, the FDIC must handle failed banks in a manner that represents the least cost to the deposit insurance fund. The only exception is if the FDIC's board of directors, the Board of Governors of the Federal Reserve System, and the secretary of the Treasury (after consultation with the president) all choose to make a systemic-risk determination, an authority that had never been used since Congress had granted it in 1991. Such a determination would be an acknowledgment that a bailout was needed to prevent significant damage to the broader financial and economic system. Such a decision also would be saying that WAMU was too big to fail.

The FDIC did not believe that WAMU was too big to fail. Our preference was protecting WAMU's uninsured depositors against any losses without violating the least-cost test. We did not care to bail out WAMU's bondholders, since they were sophisticated investors who understood (and had been well compensated for) the risks they were taking.

Jim Wigand, who managed the FDIC's resolution process for failed banks, devised a resolution strategy for WAMU. Jim had deep expertise, having worked at the RTC during the early 1990s, resolving failed banks and their assets. He had continued in a similar role at the FDIC after the two agencies merged in 1995. Jim proposed that potential acquirers would be told they had to acquire all of WAMU's assets. This would reduce the FDIC's cash outlay and relieve the agency of the costly and time-consuming task of separately having to sell $300 billion in assets. However, bidders would be given greater flexibility in determining which of WAMU's liabilities they would be willing to absorb. Their options essentially would be to assume all of the bank's liabilities less (1) all preferred stock; (2) all preferred stock and subordinated notes; (3) all preferred stock, subordinated notes, and senior debt; (4) all preferred stock, subordinated notes, senior debt, and certain liabilities; or (5) all preferred stock, subordinated notes, senior debt, certain liabilities, and uninsured deposits.

The liabilities assumed decreased from the first to the fifth option. Any liabilities assumed by the winning bidder would be protected against any losses—costs that would have to be absorbed by the acquirer. Any liabilities the winning bidder did not assume responsibility for would be exposed to losses, and represented costs the acquirer would not have to cover.

Prospective bidders were left with a difficult decision. Protecting a larger group of creditors against losses would have the desirable outcome of calming those customers and making the transaction less disruptive. However, it also would raise the cost of the transaction for the winner.

Jim had worked through the weekend of September 20 devising the strategy, and when we discussed it Sunday evening, I told him that I thought it was both elegant and straightforward. (You really have to do this for a living to think that a bidding structure for a failed bank can be elegant.) Jim had reasoned that no one would want to choose categories 1 or 2, as they would not want to absorb the costs associated with protecting WAMU's preferred shareholders and senior and subordinated debt holders. This was fine from our point of view since these private-sector creditors, rather than the FDIC's deposit insurance fund, should be first in line (after common equity shareholders) to bear the cost of the bank's failure. Jim also had reasoned that an acquirer would not want to choose categories 4 or 5 since they would not want the disruption associated with failing to protect uninsured depositors and other general creditors. This left category 3, which fit with what we wanted as well, since we had no desire to see another IndyMac-like run from depositors at WAMU's numerous branches around the country.

Early on the morning of Monday, September 22, Wigand and two other staff members took the train from Washington, D.C., to New York City to meet with prospective bidders. Citigroup, Wells Fargo, JPMC, and Santander were on their list. We planned to take bids on Wednesday and close the bank on Friday. But word of the imminent closing leaked Wednesday evening, so we decided to close the bank Thursday evening.

JPMC was the winning bidder and paid the FDIC $1.8 billion to acquire the bank. As expected, they chose to exclude from the transaction the preferred stock, subordinated debt, and senior debt. These groups would be left unprotected and have to absorb a large part of the cost associated with the bank's failure. JPMC assumed the obligation to all depositors and other creditors.

Hank Paulson and Tim Geithner later stated in their books that closing Washington Mutual and imposing losses on bondholders was a mistake. I disagree. Even if closing WAMU and imposing losses on bondholders contributed to Wachovia's liquidity problems, which it likely did indirectly, the right balance has to be struck between two competing objectives, ending an immediate crisis and not sowing the seeds for future crises. If one follows Paulson and Geithner's line of reasoning, then everyone always gets bailed out and we never put the right long-term incentives in place. Such action virtually ensures there will be more frequent and larger financial crises over time. Market discipline is essential to capitalism.

To me, the closing of WAMU was a textbook example of how a large insolvent financial institution should be handled. Despite being the largest S&L and the sixth-largest bank in the country, the closing was handled smoothly. The bank reopened for business the following day under new ownership. The closing did not cost the FDIC or taxpayers anything. Instead, the approximately $50 billion in costs were borne entirely by the private sector. JPMC wrote down the assets it acquired by about $30 billion, reflecting the unrecognized losses in WAMU's assets. The subordinated and senior debt holders lost about $11 billion of the money they had invested in the bank.

The stock market reacted positively to the news, and the next day the Dow Jones Industrial Average rose 121 points. At the FDIC, we breathed a sigh of relief and looked forward to a quiet weekend. Jim headed to the beach to relax.

Our relief lasted less than a day. Late the next afternoon, on Friday, September 26, the Treasury Department and the Federal Reserve informed the FDIC that “there may be a problem” at the Charlotte, North Carolina–based Wachovia Corporation. It appeared likely that the bank would not have enough money to reopen for business the following Monday morning. The WAMU transaction had shaken Wachovia's investors and depositors; the bank's stock price dropped 27 percent that day, and it experienced $5 billion in deposit withdrawals.

Wachovia was the fourth-largest commercial bank in the country, with over $800 billion in total assets. It operated in 21 states and in 40 countries, but its continued growth had been hampered by some awful investment decisions. The worst decision was made at the peak of the housing market, in 2006, when it bought Golden West, which owned World Savings, the second-largest S&L in the country. Golden West was a San Francisco–based S&L that had been run by Herb and Marion Sandler since they had purchased it as a small institution in 1963. Option ARMs, or pick-a-pay loans, as they called them, were their specialty. A majority of these risky loans were made in California, Florida, and Arizona—markets among the hardest hit by the subsequent housing downturn. Our plans for a peaceful weekend were over, and Wigand was summoned to return from the beach.

On Saturday morning, the FDIC was part of a conference call with the senior staff at the other bank regulatory agencies. During that call, it seemed to us that Geithner, who was head of the Federal Reserve Bank of New York at the time, wanted to see Citigroup purchase Wachovia with some government financial assistance. Our impression was that he wanted to solve two problems at the same time, since Citigroup also was struggling. My FDIC colleagues and I were skeptical, since normally we don't want to merge a failing bank into another financially strapped bank. But by Saturday afternoon, it appeared that Wells Fargo was ready to purchase Wachovia without any financial assistance from the FDIC. We went home that evening feeling hopeful that the situation was about to be resolved. However, Sunday morning that deal fell apart. It was time to scramble again.

More than 20 FDIC staff members came into the Washington, D.C., headquarters building Sunday morning to begin working. We had only a few options. Bank of America had already purchased Countrywide and Merrill Lynch. JPMC had already purchased Bear Stearns and Washington Mutual. Neither of those two institutions was in a position to make an additional acquisition. That left two potential buyers: Citigroup (despite our hesitation) and Wells Fargo. Both were interested in purchasing Wachovia on an open-bank basis if the FDIC was willing to provide enough financial assistance.

Sheila Bair wanted to explore all options, including closing the bank. This led to some concern at the Treasury, the Fed, and the Office of the Comptroller of the Currency (OCC) that perhaps the FDIC didn't understand the seriousness of the situation. They did not believe the system could withstand an additional large-bank closing, with additional losses imposed on bank bondholders.

The FDIC career staff never viewed closing Wachovia as a viable option. The problems in the financial sector had grown too deep, and Wachovia was too large, too complex, and too integrated with its affiliates and the financial system. We also had been preparing for a large-bank failure since 2004. In the years before the crisis, we had conducted simulations of large-bank failures, including one that had characteristics remarkably similar to Wachovia. The staff understood its capabilities under the resolution framework in place at that time. While we were prepared for almost any type of bank failure, there still were some banks that were too large and too interconnected for us to handle without resorting to a creditor bailout. Wachovia was one of those banks.

By Sunday night, we were using my office as the base for our discussions. It was my job to work with the FDIC's senior staff to recommend the best option to Bair and the FDIC's board of directors. Marty Gruenberg, the FDIC's vice chairman, waited with Bair as we negotiated with Citigroup and Wells Fargo. I brought Bair and Gruenberg updates about every hour. In turn, Bair would update Tim Geithner, Federal Reserve Chairman Ben Bernanke, and John Dugan, the comptroller of the currency.

We broke the staff into two teams. I led a group that began negotiations with senior representatives from Citigroup, while Wigand led a group that began similar discussions with representatives from Wells Fargo. The discussions were difficult and time consuming. We wanted to negotiate the best deal possible, but the Federal Reserve and Treasury were anxious that we not let too much time pass before making a decision.

Around midnight, Wachovia presented us a proposal to remain as a stand-alone entity with FDIC financial support. However, the plan did not look well developed and we weren't really interested in contributing money directly to Wachovia. We wanted some other group to take it over. Wachovia's proposal never received serious consideration.

The hours passed quickly. Delivery of a dozen or so pizzas was a welcome development. By early morning, we had received two very different proposals. Wells Fargo wanted the FDIC to cover most of the losses on a pool of Wachovia's assets, up to a maximum of $20 billion. This meant that the FDIC almost certainly would face a loss in the billions of dollars, but that our overall exposure would be capped at $20 billion.

Citigroup was willing to absorb the first $42 billion in losses, but they wanted the FDIC to cover any losses in excess of that amount. They were willing to pay a guarantee fee of $12 billion for that protection. Thus, the FDIC would have an open-ended exposure above $42 billion in losses, but there was a strong likelihood that Wachovia's losses would never reach that amount. It was the opposite type of structure from what Wells Fargo had proposed. In our view, we would be able to complete a transaction with Citigroup at no cost to the FDIC. This was appealing despite the unnecessary comment made by one of the Citi's senior officials that their proposal had the added attraction of “making the federal government look smart.”

The Citigroup bid represented the lowest-cost transaction for the FDIC, so at about 6 A.M. Monday, September 29, we recommended it to the FDIC board of directors. They had to meet to ratify the transaction since it involved the FDIC's providing assistance to an open bank. Bair and Gruenberg had remained in Bair's office throughout the night and attended in person. The other three board members—John Dugan, John Reich, and Tom Curry—participated by phone. The sale of Wachovia to Citigroup was a step closer.

It was the first time the FDIC's board of directors had invoked a systemic risk determination since Congress had granted the authority in 1991. The board of directors of the Federal Reserve also approved the transaction. We had worked for months with the Federal Reserve on detailed protocol on how the two organizations would analyze and ratify a systemic risk determination should one ever become necessary. That protocol never saw the light of day as the urgency of the moment precluded all else.

The visual appearance of the FDIC staff was a jarring contrast to the historic nature of the FDIC board meeting. We were a bedraggled group sitting in the FDIC's stately boardroom, with its large mahogany table. Having spent the night in the office, the staff members were dressed in shorts, jeans, and other casual wear. One person cobbled together a snappy outfit that consisted of shorts, sneakers, a T-shirt, and a sports coat.

Immediately following the board meeting, we began working to finalize the agreement with Citigroup and Wachovia. Since we had not actually closed Wachovia, the open-bank transaction needed to be approved by the Federal Reserve Board, OCC, Citigroup, and Wachovia's board of directors. The latter group was unhappy with Citigroup's $1 per share offer, but it appeared they had no alternative but to agree to the deal. However, Citigroup's representatives were slow in finalizing the detailed terms of the transaction with us. A few days passed and the term sheets for the deal still had not been signed.

Then the unexpected occurred. On Thursday, October 2, Wells Fargo presented Wachovia's board of directors with a new offer that would pay them $7 a share and would require no financial support from the FDIC. Wells Fargo had received a favorable tax ruling it had been waiting for. Wachovia's board readily accepted the new offer. The FDIC did not need to agree to the deal since it was no longer expected to provide any financial assistance.

Citigroup was livid. They thought they had a deal, and now it was gone. Thursday night, an angry Vikram Pandit, Citigroup's CEO, called Bair to complain. Geithner wasn't too happy over this turn of events either. They seemed to believe that the FDIC should have forced the earlier deal to go through. While we certainly liked the idea that under the new proposal the FDIC would not have to provide any financial assistance, the truth of the matter was that in the absence of any financial assistance on our part, the FDIC had no legal authority to force the initial deal or tell Wachovia's shareholders to whom they should sell their bank.

By November, the situation at Citigroup was precarious. The market was losing confidence in the company, which led to severe liquidity problems. Few, if any, banks better fit the definition of too big to fail. Citigroup had over $1.3 trillion in total assets and operated in more than 160 countries and jurisdictions. There was no way it could be closed without causing a global financial crisis. Intense conversations began among officials at the Treasury, the Federal Reserve, the OCC, and the FDIC about how to handle the situation.

The agencies looked at options for protecting Citigroup against the risks posed by some of its bad assets. The plan that was developed was to guarantee Citigroup against any losses above a certain threshold. A pool of $306 billion in distressed assets was identified, for which Citigroup would cover the first $37 billion in losses (a rough estimate of what the expected losses would be on those assets). The bank agreed to pay the government a fee for loss protection exceeding that amount. Negotiations then began among the Treasury, the Federal Reserve, and the FDIC over who would provide which portions of the guarantee and how much Citigroup should pay each agency.

As those negotiations proceeded we asked the other government agencies what conditions should be placed on Citigroup in order for it to receive a government guarantee on a portion of its assets. We weren't prepared for the response that came back from the Treasury and Federal Reserve. They were adamantly opposed to imposing any substantive conditions on Citigroup. During one such conference call, with the FDIC participants sitting in my office, I asked the Treasury and Fed representatives what changes were being contemplated in the bank's management. The frustration on their end of the phone was apparent. The answer was none. I was amazed at what I was hearing.

In the early 1990s, Congress placed tough restrictions on the FDIC's use of open-bank financial assistance, requiring that it be the least-costly option available, that the receiving institution's shareholders not benefit from the financial assistance, and that the institution's senior managers be replaced. A systemic-risk exception was needed if the FDIC were to override any of the restrictions on providing open-bank financial assistance. A systemic-risk exception was authorized by the FDIC, the Federal Reserve, and the Treasury, but that didn't mean we should not at least have discussed whether management at financial institutions that required bailouts should pay a price. Restrictions were eventually placed on Citigroup's executive compensation and on the payment of dividends to shareholders, but these were relatively minor and seemed to be more than what the Treasury and the Federal Reserve wanted.

The Citigroup deal was finalized on November 23, 2008. The Treasury agreed to cover the first $5 billion in any losses over the $37 billion covered by the bank. In return, the Treasury received preferred stock worth $4 billion. The FDIC provided the next $10 billion in loss protection, receiving $3 billion in preferred stock in return. The Federal Reserve agreed to absorb any additional losses on the pool of assets at no cost, in the unlikely event total losses exceeded $52 billion.

In late December 2008, the Treasury decided to support the automobile industry. General Motors was provided with up to $14.4 billion in loans and $5 billion was invested in GMAC, the company's finance subsidiary. Lending commitments were also made to Chrysler Corporation. I question the need for these bailouts, since it's much easier for large commercial firms (than large financial firms) to be kept operational and reorganized through bankruptcy proceedings, and without taxpayer assistance.

In January 2009, Bank of America, with $1.9 trillion in assets, was granted a bailout similar to Citigroup's. For a fee, the Treasury, the FDIC, and the Federal Reserve guaranteed about $100 billion of the bank's poor-quality assets.

The Treasury and the Federal Reserve drove the Citigroup and Bank of America transactions. The FDIC had resolution authority for the commercial banks within each organization, but not for their parent companies and other nonbank affiliates. The FDIC needed to approve the transactions, but was brought into the discussions late in the process and mostly followed the lead of the other two agencies.

Citigroup and Bank of America were the last of the individual bank bailouts. The banks that became insolvent afterward were small, and the FDIC used its normal resolution process to shut them down and whenever possible sell them to healthy banks. During the five-year period between January 1, 2008, and December 31, 2012, nearly 500 small banks were closed. No one was bailed out. Nevertheless, almost no one lost money other than the bank owners and the FDIC. Depositors at these banks had learned to stay safely within the $250,000 deposit insurance limit.

The FDIC's decisions to close WAMU and provide open-bank assistance for Wachovia followed the pattern set in the early 1980s with the closing of Penn Square and the subsequent decision to provide Continental Illinois with open-bank assistance. In both instances, the FDIC sought to avoid bailouts as long as feasible, thereby preserving the appropriate long-term market incentives. Only when the imploding institutions truly were too big to fail and/or the stability of the financial system was at risk did the FDIC resort to bailouts.

Subsequently, as part of the 2010 Dodd-Frank Act, Congress gave the FDIC authority to handle the resolution of virtually all systemically important financial institutions. The ability to form bridge financial institutions, provide emergency liquidity support, and leave existing U.S. contracts intact were important powers that Congress authorized as part of the FDIC's new authority. Those powers match the FDIC's preexisting powers for the resolution of insured institutions and will help reduce future bailouts.

Nevertheless, even with these powers, the FDIC would have been unlikely to have acted much differently than the Treasury and the Federal Reserve in many of the individual situations involving large financial institutions without additional changes to the financial system. The Lehman bankruptcy probably would have been handled more smoothly if the FDIC had the power to keep existing contracts within the United States in place and a source of liquidity. However, Fannie Mae, Freddie Mac, AIG, Citigroup, and Bank of America were too large, too complex, with too many unknown interconnections with other large financial institutions, and in some cases, too many international operations that were under differing and often poorly prepared resolution regimes. These factors created systemic risks that go beyond the authorities and capabilities of the FDIC's resolution framework to deal with them without resorting to some form of bailouts.

Going forward, the operational capability of bank regulators to close large, systemically important financial institutions needs improvement. Such closures must be carried out in a manner that makes the institution's creditors absorb the losses, rather than taxpayers or other third parties. Resolution authorities across the world need to better develop and coordinate their legal structures and resolution mechanisms. Structural changes are needed in the financial sector to require that financial interdependencies among large institutions are made more transparent and are better controlled. That work is under way, but not yet where it needs to be.

Large commercial banks and investment banks also should be required to hold more long-term debt. Unlike short-term debt, long-term debt cannot run when problems develop. It also can cover losses and recapitalize insolvent financial institutions. Credible resolution strategies need to be developed for the largest financial institutions and clearly articulated publicly, making it clear that long-term debt holders, rather than short-term creditors, will pay the costs associated with large financial institution failures. Only then will short-term creditors begin to feel safe enough to remain calm and not run when trouble occurs. Only then can contagion in the financial sector be controlled. Only then will we be able to end bailouts and the too-big-to-fail problem once and for all.