Financial Institutions: Causes and Consequences of Recent Bank Failures

What GAO Found

Ten states concentrated in the western, midwestern, and southeastern United States—all areas where the housing market had experienced strong growth in the prior decade—experienced 10 or more commercial bank or thrift (bank) failures between 2008 and 2011. The failures of the smaller banks (those with less than $1 billion in assets) in these states were largely driven by credit losses on commercial real estate (CRE) loans. The failed banks also had often pursued aggressive growth strategies using nontraditional, riskier funding sources and exhibited weak underwriting and credit administration practices. The rapid growth of CRE portfolios led to high concentrations that increased the banks’ exposure to the sustained real estate and economic downturn that began in 2007. GAO’s econometric model revealed that CRE concentrations and the use of brokered deposits, a funding source carrying higher risk than core deposits, were associated with an increased likelihood of failure for banks across all states during the period. Several state regulatory and community banking association officials told GAO that in some cases, the losses failed banks incurred on their CRE loans were caused by declines in the value of the underlying collateral of impaired, collateral-dependent loans. However, data are not publicly available that indicate the extent to which loan losses were driven by such declines in collateral values. Fair value accounting also has been cited as a potential contributor to bank failures, but between 2007 and 2011 fair value accounting losses in general did not appear to be a major contributor, as over two-thirds of small failed banks’ assets were not subject to fair value accounting. The Department of the Treasury and the Financial Stability Forum’s Working Group on Loss Provisioning have observed that the current accounting model for estimating credit losses is based on historical loss rates, which were low in the prefinancial crisis years. They said that earlier recognition of loan losses could have potentially lessened the impact of the crisis, when banks had to recognize the losses through a sudden series of provisions to the loan loss allowance, thus reducing earnings and regulatory capital. The Financial Accounting Standards Board has issued a proposal for public comment for a loan loss provisioning model that is more forward-looking and focuses on expected losses, which would result in banks establishing earlier recognition of loan losses for the loans they underwrite and could incentivize prudent risk management practices. Moreover, it should help address the cycle of losses and failures that emerged in the recent crisis as banks were forced to increase loan loss allowances and raise capital when they were least able to do so.

The Federal Deposit Insurance Corporation (FDIC) used shared loss agreements to help resolve failed banks at the least cost during the recent financial crisis. Under a shared loss agreement, FDIC absorbs a portion of the loss on specified assets of a failed bank that are purchased by an acquiring bank. FDIC officials, state bank regulators, community banking associations, and acquiring banks of failed institutions GAO interviewed said that shared loss agreements helped to attract potential bidders for failed banks during the financial crisis. Bank officials that acquired failed banks confirmed that they would not have purchased them without FDIC’s shared loss agreements because of uncertainty of the market and valuation of assets. FDIC said the benefits of shared loss agreements included reductions in its immediate cash needs, less disruption to failed bank customers, and the movement of assets quickly into the private sector. During 2008-2011, FDIC resolved 281 of 414 failures using shared loss agreements on assets purchased by the acquiring bank. As of December 31, 2011, Deposit Insurance Fund (DIF) receiverships made shared loss payments totaling $16.2 billion. In addition, DIF receiverships are estimated to pay an additional $26.6 billion over the duration of the shared loss agreements, resulting in total estimated lifetime losses of $42.8 billion. By comparing the estimated cost of the shared loss agreements to the estimated cost of directly liquidating the failed banks’ assets, FDIC estimates that the use of shared loss agreements saved the DIF over $40 billion. While total estimated lifetime losses of the shared loss agreements may not change, the timing of the losses may change and payments from shared loss agreements may increase as the terms of the agreements mature. FDIC officials stated that the acquiring banks are monitored for compliance with the terms and conditions of the shared loss agreements. FDIC is issuing guidance to the acquiring banks reminding them of these terms to prevent increased shared loss payments as these agreements approach maturity.

The acquisitions of failed banks by healthy banks appears to have mitigated the potentially negative effects of bank failures on communities, although the focus of local lending and philanthropy may have shifted. First, while bank failures and failed bank acquisitions can have an impact on market concentration—an indicator of the extent to which banks in the market can exercise market power, such as raising prices or reducing availability of some products and services—GAO found only a limited number of metropolitan areas and rural counties were likely to have become significantly more concentrated. The lack of increases in concentration was because in many instances, the failed banks were acquired by out-of-market institutions. Second, GAO’s econometric analysis of call report data from 2006 through 2011 found that failing small banks extended progressively less net credit as they approached failure, and that acquiring banks generally increased net credit after the acquisition. However, acquiring bank and existing peer bank officials GAO interviewed noted that in the wake of the bank failures, underwriting standards had tightened and thus credit was generally more available for small business owners who had good credit histories and strong financials than those that did not. Third, officials from regulators, banking associations, and banks GAO spoke with said that involvement in local philanthropy declined as small banks approached failure but generally increased after acquisition. Yet, these acquiring banks may not focus on the same philanthropic activities as did the failed banks. Finally, GAO econometrically analyzed the relationships among bank failures, income, unemployment, and real estate prices for all states and the District of Columbia (states) for the 1994 through 2011 period and found that bank failures in a state were more likely to affect its real estate sector than its labor market or broader economy.