FDIC Bank Failure Data: Every US Bank That Has Failed Since 1934

The Federal Deposit Insurance Corporation maintains a public list of every US bank and thrift that has failed since its founding in 1933. The list covers more than 4,000 institutions across nine decades, recording the institution name, location, charter class, failure date, acquiring institution, transaction type, total assets, total deposits, and the estimated resolution cost to the deposit insurance fund. It is one of the most complete longitudinal records of financial institution failure published by any government anywhere in the world, and it is available for bulk download at no cost.

The failure list is only part of what the FDIC publishes. Quarterly call reports— balance sheet and income statement data for every insured institution—are also publicly available through the FDIC BankFind API, giving analysts the financial inputs needed to build early-warning indicators before a failure occurs. Together, the failure list and call report data form a complete system: the outcome record and the feature set for predicting it.

The failure dataset schema

Each record in the FDIC bank failure list contains the following fields. Understanding the schema is necessary before querying the data or interpreting aggregate statistics.

• Institution name and location. The legal name of the failed institution at the time of failure, the city, and the state. The name reflects the institution as it existed at failure—not any predecessor name from mergers or prior charters.
• Certificate number (cert). The FDIC certificate number is the primary identifier for insured depository institutions. It is stable across the institution's life and is the join key to call report data in the FDIC Statistics on Depository Institutions (SDI) database.
• Charter class (classcode). Identifies the regulator and charter type. Values include: N (national bank, OCC charter), SM (state member bank, Federal Reserve supervision),NM (state non-member bank, FDIC supervision), SB (savings bank), SA (savings association, OTS/OCC charter), and OI(insured US branch of a foreign bank). Charter class determines which federal agency was the primary regulator at the time of failure.
• Failure date (faildate). The date the institution was closed by its chartering authority and the FDIC was appointed receiver. For bank failures, this is almost always a Friday evening—regulators close banks at the end of business on Friday to allow the weekend for resolution operations before markets reopen Monday.
• Fund (savr). The deposit insurance fund that bore the resolution cost. Values are BIF (Bank Insurance Fund),SAIF (Savings Association Insurance Fund), and DIF (Deposit Insurance Fund). BIF and SAIF existed as separate funds until the Federal Deposit Insurance Reform Act of 2005 merged them into the DIF effective 2006. Failures before 2006 carry either BIF or SAIF depending on the institution's charter type; all failures from 2006 onward carry DIF.
• Transaction type (restype). How the FDIC resolved the failure. The main transaction types are:
  • P&A — Purchase and Assumption. An acquiring institution purchases some or all assets and assumes some or all deposits. The most common resolution type. Depositors are transferred to the acquirer with no interruption.
  • IDT — Insured Deposit Transfer. The FDIC transfers insured deposits to an acquirer without a sale of assets. Used when no buyer for the assets can be found quickly.
  • OA — Open Bank Assistance. Direct financial assistance to a troubled institution to keep it open. Rare and required a systemic risk determination; used in the 1980s S&L crisis and briefly discussed during the 2023 failures.
  • PA — Payoff. The FDIC pays insured depositors directly and liquidates the bank's assets. Used when no buyer can be found and the institution is too small to warrant a bridge bank.
  • OAA — Open Bank Assistance with Acquisition. A hybrid: assistance provided to facilitate acquisition by a healthy institution.
  • OAR — Open Assistance with Reorganization. Assistance conditioned on reorganization, used primarily in the thrift crisis.
• Total assets and total deposits. Measured in thousands of dollars as of the date of failure. These are book values from the institution's last call report filing, not liquidation values.
• Resolution cost (cost). The FDIC's estimated cost to the deposit insurance fund, in thousands of dollars. This is a net present value estimate that includes projected recoveries from asset liquidation. Final cost figures are revised as asset portfolios are worked out, sometimes over many years. The figures in the failure list reflect the most recent FDIC estimate.
• Acquiring institution (acqinst). The institution that assumed deposits or purchased assets in a P&A transaction. “No Acquirer” appears for payoff resolutions.

Historical failure waves

The 4,000+ failures recorded since 1934 are not distributed evenly across time. They cluster into three distinct waves, each driven by a different macroeconomic shock and regulatory failure.

The S&L Crisis (1980–1994)

The largest wave in the FDIC's history by institution count. More than 3,000 banks and thrifts failed between 1980 and 1994—roughly 75% of all failures since 1934 concentrated in a 15-year span. The crisis had two intertwined causes. Interest rate deregulation under the Depository Institutions Deregulation and Monetary Control Act of 1980 allowed thrifts to pay market rates on deposits while remaining locked into long-duration fixed-rate mortgage portfolios from earlier decades. The mismatch between short-term funding costs and long-term asset yields devastated thrift balance sheets as the Volcker Fed drove interest rates above 20% in 1981.

The second phase of the crisis was driven by real estate speculation and outright fraud. Deregulation allowed thrifts to invest in commercial real estate and junk bonds, asset classes that collapsed in the late 1980s. The Lincoln Savings failure—Charles Keating's institution, closed in 1989 with a resolution cost of approximately $3 billion—became the symbol of the era. The Resolution Trust Corporation (RTC) was created in 1989 specifically to manage failed thrift assets; it handled roughly 750 institution failures before being folded into the FDIC in 1995.

The GFC wave (2008–2012)

The global financial crisis produced 500+ bank failures between 2008 and 2012, peaking at 157 failures in 2010. The failure profile was different from the S&L crisis: most GFC-era failures were community banks with concentrated exposure to construction and land development loans in overbuilt real estate markets—particularly in Georgia, Florida, Illinois, and California. Washington Mutual's failure in September 2008— $307 billion in assets—remains the largest bank failure in US history by assets. IndyMac (closed July 2008, $32 billion in assets) was the largest FDIC-managed failure until WaMu's collapse two months later.

The GFC wave differed from the S&L crisis in resolution mechanics. Most GFC-era community bank failures were resolved via P&A transactions with regional bank acquirers, limiting depositor disruption. The resolution cost as a percentage of assets was also lower than the thrift crisis, partly because the FDIC had developed more sophisticated loss-sharing agreements that transferred credit risk to acquirers while limiting the acquirer's downside on the most impaired assets.

The 2023 episode

Three failures in the spring of 2023 put more assets into FDIC receivership in a single year than any year since 2008—and by some measures, the most in history on a dollar basis. Silicon Valley Bank failed on March 10, 2023, with $209 billion in assets and $175 billion in deposits. Signature Bank failed two days later with $110 billion in assets. First Republic Bank failed on May 1, 2023, with $213 billion in assets, making it the second-largest bank failure in US history.

All three failures shared a structural feature: a deposit base dominated by accounts far above the $250,000 insurance limit, held by venture capital firms, tech companies, and cryptocurrency businesses. SVB's uninsured deposits were approximately 94% of total deposits at the time of failure. The concentration of uninsured depositors in institutions that held long-duration securities portfolios funded by short-term deposits created a run dynamic identical in structure—if not in cause—to the thrift crisis's duration mismatch problem.

The FDIC invoked the systemic risk exception under the Federal Deposit Insurance Act to guarantee all deposits at SVB and Signature—including amounts above the $250,000 limit—without requiring Congressional authorization. This exception, last used during the GFC in 2008, requires a two-thirds vote of the FDIC board, a two-thirds vote of the Federal Reserve Board, and the written recommendation of the Treasury Secretary.

Deposit insurance limit history

The federal deposit insurance limit has been raised four times since the FDIC was founded. Understanding the limit history is essential for interpreting the failure record, because the insured versus uninsured deposit split determines both the depositor harm from a failure and the FDIC's resolution strategy.

• 1934–1950: $5,000 per depositor.
• 1950–1966: $10,000 per depositor.
• 1966–1969: $15,000 per depositor.
• 1969–1974: $20,000 per depositor.
• 1974–1980: $40,000 per depositor.
• 1980–2008: $100,000 per depositor. The limit was set at $100,000 by the Depository Institutions Deregulation and Monetary Control Act of 1980 and remained there for 28 years.
• 2008–present: $250,000 per depositor. Raised temporarily by the Emergency Economic Stabilization Act of 2008 and made permanent by the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010.

The limit applies per depositor, per institution, per ownership category. A depositor with individual accounts, joint accounts, and IRA accounts at the same institution can hold more than $250,000 in total insured coverage because each ownership category is separately insured.

BIF, SAIF, and the Deposit Insurance Fund

Before 2006, the FDIC administered two separate deposit insurance funds with different histories, different premium bases, and different capital adequacy situations. The distinction appears throughout the failure list and matters for any pre-2006 analysis.

The Bank Insurance Fund (BIF) covered commercial banks and savings banks that elected BIF coverage. It was recapitalized after the S&L crisis through mandatory premium assessments and reached its statutory minimum reserve ratio of 1.25% of insured deposits in 1995. Following recapitalization, Congress reduced BIF premiums to near zero for well-capitalized institutions, which created a perverse situation: thousands of healthy banks paid essentially nothing into the fund they were drawing on for failure resolution.

The Savings Association Insurance Fund (SAIF)covered savings institutions—thrifts, savings banks, and savings associations—and inherited the liabilities of the now-defunct FSLIC (Federal Savings and Loan Insurance Corporation), which had been insolvent during the thrift crisis and was abolished in 1989. SAIF was undercapitalized relative to BIF through most of the 1990s because the thrift industry it covered remained troubled long after the commercial banking sector recovered. Congress imposed a one-time SAIF premium assessment in 1996 to recapitalize the fund.

The Federal Deposit Insurance Reform Act of 2005 merged BIF and SAIF into the single Deposit Insurance Fund (DIF) effective January 1, 2006. The DIF is funded through risk-based premiums set by the FDIC under a formula that accounts for an institution's capital ratios, supervisory ratings, and other risk indicators. The Dodd-Frank Act raised the DIF minimum reserve ratio from 1.15% to 1.35% of estimated insured deposits and required the FDIC to achieve that ratio by September 30, 2020.

Call report data and the Texas Ratio

The failure list records outcomes. The FDIC's call report data provides the inputs for predicting them. Every FDIC-insured institution files quarterly call reports— formally the FFIEC 031 and FFIEC 041—containing a complete balance sheet, income statement, and detailed schedules on loan portfolio composition, asset quality, capital adequacy, and liquidity. The data is available through the FDIC Statistics on Depository Institutions (SDI) database and the FDIC BankFind API going back to the 1990s.

The most widely used single-metric failure predictor derived from call report data is the Texas Ratio, developed by Gerard Cassidy at RBC Capital Markets during the Texas real estate collapse of the late 1980s. The formula is:

Texas Ratio = Non-Performing Assets / (Tangible Common Equity + Loan Loss Reserves)

Where:
  Non-Performing Assets = Non-accrual loans
                        + Loans 90+ days past due
                        + Other Real Estate Owned (OREO)
  Tangible Common Equity = Total equity capital - Intangible assets - Goodwill
  Loan Loss Reserves = Allowance for Loan and Lease Losses (ALLL)

A Texas Ratio above 100% means the bank's non-performing assets exceed its capacity to absorb losses from tangible equity and loan loss reserves combined. At that point, any further deterioration in the loan portfolio is likely to impair insured deposits and require FDIC intervention. Empirical research on the GFC-era failure wave found that institutions with a Texas Ratio above 100% failed at rates exceeding 85% in the subsequent 12 months, compared to a base failure rate below 1% for the overall insured institution population.

The Texas Ratio can be constructed entirely from publicly available FDIC call report data. The relevant call report schedule fields are: Schedule RC-N (past-due and non-accrual loans), Schedule RC-M (OREO), Schedule RC (total equity and intangibles), and Schedule RI-B (allowance for loan losses). All are available quarterly through the FDIC BankFind API.

Accessing the data: FDIC BankFind Suite

The primary access point for FDIC data is the FDIC BankFind Suite at banks.fdic.gov. The suite provides three access methods for the failure list and call report data:

• BankFind API. A REST API at banks.fdic.gov/api/ with endpoints for failures (/failures), institutions (/institutions), financial call report data (/financials), and history records (/history). The API supports field selection, filtering, sorting, and pagination. Output formats include JSON and CSV. No API key is required.
• Bulk CSV download. The complete failure list and institution list are available as direct CSV downloads from the BankFind Suite. The failure list CSV is updated within days of each new failure and contains the full history from 1934 onward. For one-off analysis, this is the fastest path.
• FDIC Statistics on Depository Institutions (SDI).The SDI tool at banks.fdic.gov/sdi/ provides a web interface for querying call report data across the full institution population. Custom peer group analysis, time-series charts, and multi-institution comparisons are available through the SDI interface. The underlying data is also accessible via the BankFind API's/financials endpoint.

Python: downloading and analyzing by era

The following script downloads the complete FDIC failure list via the BankFind API, assigns era labels, and computes state-level failure counts and aggregate resolution statistics by era. The qbfasset and qbfdep fields are in thousands of dollars; the script converts to billions for readability.

import requests
import pandas as pd

# Download the FDIC bank failure list as CSV
FAILURES_URL = 'https://banks.fdic.gov/api/failures?fields=name,city,stname,cert,savr,restype,resdate,faildate,qbfasset,qbfdep,cost,acqinst&limit=10000&offset=0&output=csv'

df = pd.read_csv(FAILURES_URL)
df['faildate'] = pd.to_datetime(df['faildate'], errors='coerce')
df['year'] = df['faildate'].dt.year

# Assign era labels
def era(y):
    if pd.isna(y):
        return 'Unknown'
    if y < 1980:
        return 'Pre-S&L (1934-1979)'
    if y <= 1994:
        return 'S&L Crisis (1980-1994)'
    if y <= 2007:
        return 'Quiet period (1995-2007)'
    if y <= 2012:
        return 'GFC wave (2008-2012)'
    return 'Post-GFC (2013+)'

df['era'] = df['year'].apply(era)

# State-level failure rates by era
state_era = (
    df.groupby(['stname', 'era'])
    .size()
    .reset_index(name='failures')
    .sort_values(['stname', 'era'])
)

# Total assets at failure by era (qbfasset is in thousands)
era_summary = (
    df.groupby('era')
    .agg(
        failures=('cert', 'count'),
        total_assets_bn=('qbfasset', lambda x: x.sum() / 1_000_000),
        total_deposits_bn=('qbfdep', lambda x: x.sum() / 1_000_000),
        resolution_cost_bn=('cost', lambda x: x.sum() / 1_000_000),
    )
    .reset_index()
)

print(era_summary.to_string(index=False))
print()
print("Top 10 states by total failures:")
print(state_era.groupby('stname')['failures'].sum().nlargest(10))

The script outputs an era summary table showing failures, total assets, total deposits, and resolution costs by period. Running it against the current failure list, the S&L Crisis era (1980–1994) accounts for roughly 3,100 failures, while the GFC wave (2008–2012) accounts for approximately 500. Texas, California, Illinois, Florida, and Georgia consistently rank in the top five states by total failure count across both waves, reflecting the concentration of thrift activity and construction lending in those states.

How analysts use FDIC data to identify at-risk institutions

Financial journalists and institutional analysts building early-warning screens typically combine several call report metrics rather than relying on the Texas Ratio alone. A standard screening workflow looks for institutions satisfying multiple thresholds simultaneously:

• Texas Ratio above 50%. This is the “watch list” threshold. Institutions above 50% are not necessarily at imminent risk, but they are worth monitoring quarterly. Above 100%, the risk profile changes materially.
• Non-accrual loans as a percentage of total loans above 5%. Available from Schedule RC-N. A ratio above 5% indicates a non-trivial portion of the loan portfolio is generating no interest income.
• Tier 1 capital ratio below 6%. The regulatory minimum for “adequately capitalized” is 4%. Institutions between 4% and 6% are technically adequate but lack cushion against further credit deterioration. Prompt corrective action provisions kick in below 4%.
• Loan-to-deposit ratio above 90% combined with a high concentration in commercial real estate. Institutions that funded concentrated CRE portfolios with deposit growth are exposed to both credit risk and funding risk if deposit outflows accelerate. Call report Schedule RC-C provides the CRE concentration breakdown; the loan-to-deposit ratio is total loans divided by total deposits from Schedule RC.
• Declining net interest margin with a long-duration securities portfolio. The SVB failure pattern. When a bank holds large amounts of available-for-sale or held-to-maturity securities with long durations, rising interest rates reduce the market value of those securities. The unrealized loss position is visible in Schedule RC and is now separately disclosed following regulatory guidance issued after 2023.

The FDIC itself publishes a problem bank list each quarter in its Quarterly Banking Profile, reporting the number of institutions on its internal watch list and aggregate assets, without naming the institutions. The undisclosed “watch list” count is a leading indicator: when the number of problem institutions rises sharply, a wave of FDIC-resolvable failures typically follows 12–24 months later.

For large institution failures—the SVB/Signature/First Republic class of event—the early-warning signals were visible in public call report data months before closure. SVB's held-to-maturity securities portfolio had an unrealized loss exceeding its tangible equity as of December 31, 2022, a fact that was disclosed in its year-end 10-K. Signature Bank's uninsured deposit concentration was documented in every quarterly call report. The data existed; the analytical attention was not consistently directed at it until the failures were imminent.

For SEC Form 13F institutional holdings data—which documented which pension funds and hedge funds held SVB, Signature, and First Republic through their 2023 collapses—and how to join quarterly position data against FDIC failure records: Who owns what: indexing SEC Form 13F institutional holdings data →

For FEC campaign finance data and how financial institution PACs appear in the committee master file alongside the same institutions' FDIC call report profiles: Follow the money: mapping dark money and super PAC flows with FEC bulk data →

For PCAOB audit deficiency data—relevant to evaluating the auditors of financial institutions flagged by Texas Ratio screens, since audit quality failures and bank distress often coincide: PCAOB audit deficiency data: inspection reports, Part I findings, and firm-level risk signals →