US Bank Failures 2026
By Axis Intelligence Research
Co-author: Sarah Davis | Last updated: October 12, 2026 | License: CC BY 4.0
Six FDIC-insured banks have failed in 2026 as of October 12, holding $1.45 billion in assets and costing the Deposit Insurance Fund an estimated $243.3 million, according to Axis Intelligence Research’s count of FDIC records. On the Axis Failure Loss Severity Ratio (FLSR), this year’s failures cost the fund 16.8 cents per dollar of failed-bank assets.
Quick Answer
Six US banks have failed in 2026, the most in a single year since 2017, led by Nano Banc of Irvine, California ($736 million in assets, September 25). Since 1934 the FDIC has handled 3,526 bank failures. According to Axis Intelligence Research, failures under $100 million cost the fund 4.3 times more per dollar than failures above $10 billion.
Key Findings
- According to Axis Intelligence Research, bank failures under $100 million in assets cost the Deposit Insurance Fund 25.69 cents per dollar of assets from 2008 to 2026, against 6.03 cents for failures above $10 billion.
- Silicon Valley Bank and First Republic Bank account for $34.24 billion, or 32.0%, of every dollar of failure losses the FDIC has recorded since 2008, Axis Intelligence Research finds.
- Georgia had 92 bank failures between 2008 and October 2026, equal to 26.1% of the 352 insured institutions it had at the end of 2007, according to Axis Intelligence Research.
- FDIC resolutions that transferred only insured deposits to a buyer lost 38.78 cents per dollar of assets since 2008, more than five times the 7.39 cents on standard purchase-and-assumption deals, per Axis Intelligence Research.
- Nano Banc, with $736 million in assets, is the largest US bank failure since Republic First Bank in April 2024, and 2026’s six failures are the most in any year since 2017.
How Many Banks Have Failed in 2026?
Six FDIC-insured banks have failed in 2026 as of October 12. Four appear in the FDIC’s failed bank data on BankFind Suite; the two most recent, Tioga-Franklin Savings Bank on August 21 and Nano Banc on September 25, were announced in FDIC press releases and had not yet been added to the downloadable failure file when this dataset was compiled. We count all six.
| Bank | Location | Failure date | Assets at failure | Deposits at failure | Estimated loss to the DIF | FLSR | Source |
|---|---|---|---|---|---|---|---|
| Metropolitan Capital Bank and Trust | Chicago, IL | Jan 30, 2026 | $261.2 million | $212.2 million | $19.6 million | 7.52% | FDIC failure data |
| Community Bank and Trust – West Georgia | LaGrange, GA | May 1, 2026 | $305.7 million | $296.4 million | $97.3 million | 31.82% | FDIC failure data |
| Kentland Federal Savings and Loan Association | Kentland, IN | Jul 10, 2026 | $3.8 million | $3.7 million | $1.2 million | 31.36% | FDIC failure data |
| Small Business Bank | Lenexa, KS | Jul 17, 2026 | $72.9 million | $68.8 million | $5.7 million | 7.77% | FDIC failure data |
| Tioga-Franklin Savings Bank | Philadelphia, PA | Aug 21, 2026 | $68 million | $67 million | $5.5 million (preliminary) | 8.09% | FDIC press release |
| Nano Banc | Irvine, CA | Sep 25, 2026 | $736 million | $686 million | $114 million (preliminary) | 15.49% | FDIC press release |
| Total, 2026 | $1,447.7 million | $1,334.0 million | $243.3 million | 16.81% | Axis Intelligence Research |
Press-release assets and deposits are June 30, 2026 Call Report values as stated by the FDIC. FLSR is calculated by Axis Intelligence Research.
Nano Banc is the largest failure of the year by a wide margin. The FDIC’s announcement puts its assets at $736 million and deposits at $686 million, with Sunwest Bank of Sandy, Utah, assuming substantially all deposits and an estimated cost to the fund of about $114 million. That makes it the largest US bank failure since Republic First Bank of Philadelphia, which held $5,866.2 million in assets when it failed on April 26, 2024.
The smallest was Kentland Federal Savings and Loan Association, a $3.8 million Indiana thrift. The OCC’s closing release cites losses that depleted its capital and left it critically undercapitalized; the OCC put its assets at about $3.7 million at March 31, 2026, against $3.808 million in the FDIC file, a gap we carry in the dataset as a source discrepancy. In Philadelphia, the Pennsylvania Department of Banking and Securities took possession of Tioga-Franklin citing an unsafe and unsound condition, and the FDIC arranged for Second Federal Savings and Loan Association of Philadelphia to assume all deposits.
Which 2026 bank failures cost the FDIC the most?
Two of the six account for 86.8% of the year’s estimated losses: Nano Banc at $114 million and Community Bank and Trust – West Georgia at $97.3 million. Both exceed the $50 million line that, under 12 U.S.C. 1831o(k), defines a material loss and requires the inspector general of the bank’s federal regulator to report on why the failure cost the fund so much. According to Axis Intelligence Research, 2026 is the first year with two material-loss failures since 2023.
The LaGrange, Georgia, failure stands out for another reason. The FDIC resolved it as a purchase and assumption of insured deposits only, the transaction type the FDIC codes PI, so the buyer did not take on uninsured balances. Its FLSR of 31.82% is the highest of any 2026 failure.
The year in context: 2026’s failed assets are 11.5 times 2025’s $126.4 million, yet they equal 0.0055% of the $26.46 trillion held by US banks, and the year’s estimated losses equal 0.15% of the Deposit Insurance Fund’s $161.1 billion balance at June 30, 2026, by Axis Intelligence Research’s calculation.
Sarah Davis: Six failures sounds like a trend line, and it isn’t one yet. These are small, locally funded banks that ran out of capital one at a time, not a run spreading through a funding market. The number worth watching is the loss per dollar. At 16.8 cents, the 2026 failures are expensive for their size, which tells you supervisors are closing banks late, after the asset book has already rotted.
How Many Banks Have Failed in the US Since 1934?
The FDIC has recorded 3,526 bank failures from its first year of operation in 1934 through October 12, 2026: the 3,524 in its failure file plus the two 2026 failures announced since. The same file lists 593 separate open-bank assistance transactions, which we exclude because those banks did not close.
| Period | Failures | Assets at failure | Estimated loss to the DIF | FLSR | Average loss per failure | Source |
|---|---|---|---|---|---|---|
| 1934 to 1979 | 558 | $8.6 billion | Not recorded | n/a | n/a | FDIC failure data |
| 1980 to 1994 (thrift and banking crisis) | 2,359 | $658.7 billion | $127.9 billion | 19.41% | $54.2 million | Axis Intelligence Research from FDIC data |
| 1995 to 2007 | 58 | $12.2 billion | $2.0 billion | 16.03% | $33.7 million | Axis Intelligence Research from FDIC data |
| 2008 to 2014 (financial crisis wave) | 507 | $689.2 billion | $69.6 billion | 10.1% | $137.3 million | Axis Intelligence Research from FDIC data |
| 2015 to 2022 | 29 | $12.7 billion | $2.0 billion | 15.86% | $69.6 million | Axis Intelligence Research from FDIC data |
| 2023 | 5 | $532.2 billion | $34.3 billion | 6.44% | $6,859.0 million | Axis Intelligence Research from FDIC data |
| 2024 to 2026 (to Oct 12) | 10 | $7.5 billion | $1.0 billion | 13.77% | $104.0 million | Axis Intelligence Research from FDIC data |
The FDIC file records no loss estimates for failures before 1980, so no FLSR is computed for that period. Figures are nominal dollars.
The peak year was 1989, with 530 failures as the savings and loan collapse played out. In the Deposit Insurance Fund era, the peak came in 2010, with 157. The 2008 to 2014 wave closed 507 banks holding $689.2 billion in assets, equal to 5.29% of the $13,033.9 billion US banks held at the end of 2007, according to Axis Intelligence Research. The five failures of 2023 held assets equal to 2.26% of the industry’s balance sheet at the end of 2022. Five banks delivered a shock 42.7% the relative size of the seven-year 2008 to 2014 wave.
Bank failures by year since 2007
| Year | Failures | Assets at failure | Estimated loss to the DIF | FLSR | Source |
|---|---|---|---|---|---|
| 2007 | 3 | $2.6 billion | $157.4 million | 6.02% | FDIC failure data |
| 2008 | 25 | $371.9 billion | $17.7 billion | 4.77% | FDIC failure data |
| 2009 | 140 | $169.7 billion | $25.8 billion | 15.2% | FDIC failure data |
| 2010 | 157 | $92.1 billion | $15.8 billion | 17.12% | FDIC failure data |
| 2011 | 92 | $34.9 billion | $6.4 billion | 18.25% | FDIC failure data |
| 2012 | 51 | $11.6 billion | $2.4 billion | 20.46% | FDIC failure data |
| 2013 | 24 | $6.0 billion | $1.2 billion | 19.9% | FDIC failure data |
| 2014 | 18 | $2.9 billion | $375.1 million | 12.88% | FDIC failure data |
| 2015 | 8 | $6.7 billion | $847.0 million | 12.63% | FDIC failure data |
| 2016 | 5 | $277.2 million | $42.5 million | 15.32% | FDIC failure data |
| 2017 | 8 | $5.1 billion | $1.0 billion | 19.94% | FDIC failure data |
| 2018 | 0 | $0 | $0 | n/a | FDIC failure data |
| 2019 | 4 | $208.8 million | $25.3 million | 12.13% | FDIC failure data |
| 2020 | 4 | $455.0 million | $90.1 million | 19.8% | FDIC failure data |
| 2021 | 0 | $0 | $0 | n/a | FDIC failure data |
| 2022 | 0 | $0 | $0 | n/a | FDIC failure data |
| 2023 | 5 | $532.2 billion | $34.3 billion | 6.44% | FDIC failure data |
| 2024 | 2 | $6.0 billion | $754.7 million | 12.63% | FDIC failure data |
| 2025 | 2 | $126.4 million | $41.6 million | 32.92% | FDIC failure data |
| 2026 (to Oct 12) | 6 | $1.4 billion | $243.3 million | 16.81% | FDIC failure data and press releases |
FLSR for each year is an Axis Intelligence Research calculation. The 2023 figure records Signature Bank’s loss at zero, as the FDIC file does; see the 2023 section below.
How often does a year pass with no bank failures?
Rarely. According to Axis Intelligence Research’s count, only five calendar years since 1934 have passed without a single FDIC-insured bank failure: 2005, 2006, 2018, 2021 and 2022. Every one of them falls in the last two decades.
Sarah Davis: The cost column has a pattern the count column hides. In 2008, the worst-looking year by assets, the fund lost 4.77 cents per dollar, because Washington Mutual’s $307.0 billion failed with no recorded loss. The expensive years came later: 2012 cost 20.46 cents per dollar on banks almost nobody outside their county had heard of. A failure wave starts with headline names and ends with hundreds of small lenders that sat on bad construction loans for too long.
What Are the Largest Bank Failures in US History?
Washington Mutual Bank is the largest bank failure in US history, with $307.0 billion in assets when it failed on September 25, 2008. Three of the four largest failures happened within eight weeks in 2023.
| Rank | Bank | Location | Failure date | Assets at failure | Estimated loss to the DIF | FLSR | Source |
|---|---|---|---|---|---|---|---|
| 1 | Washington Mutual Bank | Henderson, NV | Sep 25, 2008 | $307.0 billion | $0 recorded | n/a | FDIC failure data |
| 2 | First Republic Bank | San Francisco, CA | May 1, 2023 | $212.6 billion | $15.50 billion | 7.29% | FDIC failure data |
| 3 | Silicon Valley Bank | Santa Clara, CA | Mar 10, 2023 | $209.0 billion | $18.74 billion | 8.97% | FDIC failure data |
| 4 | Signature Bank | New York, NY | Mar 12, 2023 | $110.4 billion | $0 recorded (FDIC 2023 estimate: about $2.5 billion) | n/a | FDIC failure data; FDIC press release |
| 5 | IndyMac Bank | Pasadena, CA | Jul 11, 2008 | $30.7 billion | $11.98 billion | 39.02% | FDIC failure data |
| 6 | Colonial Bank | Montgomery, AL | Aug 14, 2009 | $25.5 billion | $2.37 billion | 9.3% | FDIC failure data |
| 7 | First RepublicBank-Dallas | Dallas, TX | Jul 29, 1988 | $17.1 billion | $2.02 billion | 11.81% | FDIC failure data |
| 8 | Guaranty Bank | Austin, TX | Aug 21, 2009 | $13.5 billion | $0 recorded | n/a | FDIC failure data |
| 9 | Bank of New England | Boston, MA | Jan 6, 1991 | $13.4 billion | $0.57 billion | 4.26% | FDIC failure data |
| 10 | Gibraltar Savings | Simi Valley, CA | Mar 31, 1989 | $13.4 billion | $0.09 billion | 0.71% | FDIC failure data |
Assets are as reported on each bank’s last financial report before failure. Losses are the FDIC’s current estimates; Washington Mutual, Signature Bank and Guaranty Bank are recorded at zero in the FDIC file.
IndyMac is the expensive outlier: a $30.7 billion thrift that cost the fund $11.98 billion, or 39.02 cents per dollar, a severity rate more typical of a small community bank. Silicon Valley Bank carries the largest single recorded loss in the file at $18.74 billion.
How Much Do Bank Failures Cost the FDIC? The Failure Loss Severity Ratio (FLSR)
Counting failures tells you how many banks closed. It doesn’t tell you how badly. The Axis Failure Loss Severity Ratio (FLSR) measures what a failure actually costs the Deposit Insurance Fund relative to the size of the bank that failed.
How the FLSR is calculated
FLSR = FDIC estimated loss to the insurance fund ÷ total assets at failure × 100
Both inputs come straight from the FDIC: the loss is the FDIC’s estimate of what it disbursed to cover insured deposits minus what it expects to recover from the failed bank’s assets, as defined in the FDIC’s failure data dictionary, and assets are those on the bank’s last report before closing. For any group of failures, Axis Intelligence Research sums the losses and divides by the summed assets, so the ratio is dollar-weighted. An FLSR of 16.81% reads simply: the fund lost 16.81 cents for every dollar of assets on the failed banks’ balance sheets.
According to Axis Intelligence Research, the FLSR across all 551 failures from 2008 to October 12, 2026 is 8.61%, on $106.97 billion of losses and $1.24 trillion of failed assets. That aggregate is pulled down by the two largest failures recorded at zero loss: excluding Washington Mutual and Signature Bank, the 2008 to 2026 FLSR is 12.98%.
Why do small bank failures cost the FDIC more per dollar?
Size is the single strongest predictor of severity in the FDIC data.
| Assets at failure | Failures, 2008 to 2026 | Assets at failure | Estimated loss to the DIF | FLSR | Average loss per failure | Source |
|---|---|---|---|---|---|---|
| Under $100 million | 144 | $8.0 billion | $2.05 billion | 25.69% | $14.3 million | Axis Intelligence Research from FDIC data |
| $100 million to $1 billion | 337 | $110.5 billion | $24.69 billion | 22.34% | $73.3 million | Axis Intelligence Research from FDIC data |
| $1 billion to $10 billion | 58 | $155.5 billion | $21.83 billion | 14.04% | $376.4 million | Axis Intelligence Research from FDIC data |
| Over $10 billion | 12 | $967.7 billion | $58.40 billion | 6.03% | $4,866.6 million | Axis Intelligence Research from FDIC data |
According to Axis Intelligence Research, a failed bank under $100 million cost the fund 4.3 times as much per dollar of assets as a failed bank above $10 billion. Even after removing the three large failures recorded at zero loss (Washington Mutual, Signature Bank and Guaranty Bank), the over-$10 billion group’s FLSR is 10.88%, still well under half the small-bank rate.
The 2026 failures follow the same curve. Kentland’s $3.8 million balance sheet cost 31.36 cents per dollar; Nano Banc’s $736 million cost 15.49 cents.
Does the way the FDIC resolves a failure change the loss?
Yes, sharply. Purchase and assumption, where a healthy bank buys the failed bank and takes on its deposits, is the FDIC’s standard tool: it was used in 508 of 549 failures in the FDIC file from 2008 to July 2026, or 92.5%.
| Resolution type (FDIC code) | Failures, 2008 to 2026 | Assets at failure | Estimated loss to the DIF | FLSR | Source |
|---|---|---|---|---|---|
| Purchase and assumption of insured and uninsured deposits (PA) | 508 | $1,186.6 billion | $87.74 billion | 7.39% | Axis Intelligence Research from FDIC data |
| Purchase and assumption of insured deposits only (PI) | 15 | $38.4 billion | $14.90 billion | 38.78% | Axis Intelligence Research from FDIC data |
| Payout to depositors (PO) | 18 | $10.7 billion | $2.22 billion | 20.78% | Axis Intelligence Research from FDIC data |
| Deposit Insurance National Bank (DINB) | 8 | $5.2 billion | $1.99 billion | 38.07% | Axis Intelligence Research from FDIC data |
The FDIC file stops at the July 17, 2026 failure; resolution types for the two later 2026 failures are not yet coded.
Bank failures by charter type
State-chartered banks that are not members of the Federal Reserve, supervised by the FDIC itself, made up 332 of the 549 failures in the FDIC file since 2008, or 60.5%. Their FLSR of 10.18% is nearly identical to national banks supervised by the OCC (10.05%) and state member banks supervised by the Federal Reserve (10.22%). The six savings and loan associations in the file carry the highest severity, at 26.88%.
Sarah Davis: The size table is the most useful number in deposit insurance, and it rarely gets printed. A $60 million bank has no franchise a buyer will pay for: no deposit base worth inheriting, no fee business, just a loan book the FDIC has to mark down. A large bank, even a broken one, has branches, clients and systems someone wants. That is why the fund loses a quarter of the balance sheet on the small ones and a sixteenth on the giants. And the PI row is the same story told by the buyer: when no acquirer will take the uninsured depositors along, the assets are usually the reason.
Which States Have Had the Most Bank Failures Since 2008?
Georgia has had more bank failures than any other state since 2008, with 92, followed by Florida (73), Illinois (68) and California (43). Counts favor states with many banks, so Axis Intelligence Research also measures each state’s failures from 2008 to October 2026 against the number of FDIC-insured institutions it had at the end of 2007, from the FDIC’s historical bank data.
| Rank | State | Failures, 2008 to 2026 | Insured institutions, end of 2007 | Failures per 100 institutions | Estimated loss to the DIF | FLSR | Source |
|---|---|---|---|---|---|---|---|
| 1 | Puerto Rico | 4 | 10 | 40.0 | $4,380.2 million | 17.64% | Axis Intelligence Research from FDIC data |
| 2 | Nevada | 12 | 44 | 27.3 | $2,372.5 million | n/a | Axis Intelligence Research from FDIC data |
| 3 | Arizona | 15 | 57 | 26.3 | $505.8 million | 21.4% | Axis Intelligence Research from FDIC data |
| 4 | Georgia | 92 | 352 | 26.1 | $10,987.8 million | 32.07% | Axis Intelligence Research from FDIC data |
| 5 | Florida | 73 | 317 | 23.0 | $10,460.6 million | 26.84% | Axis Intelligence Research from FDIC data |
| 6 | Washington | 19 | 97 | 19.6 | $1,751.9 million | 15.67% | Axis Intelligence Research from FDIC data |
| 7 | Oregon | 6 | 40 | 15.0 | $256.8 million | 11.04% | Axis Intelligence Research from FDIC data |
| 8 | California | 43 | 312 | 13.8 | $50,439.3 million | 9.59% | Axis Intelligence Research from FDIC data |
| 9 | South Carolina | 10 | 93 | 10.8 | $591.2 million | 19.44% | Axis Intelligence Research from FDIC data |
| 10 | Idaho | 2 | 19 | 10.5 | $112.7 million | 17.5% | Axis Intelligence Research from FDIC data |
| 11 | Maryland | 10 | 97 | 10.3 | $408.5 million | 17.47% | Axis Intelligence Research from FDIC data |
| 12 | Utah | 7 | 68 | 10.3 | $1,037.0 million | 31.46% | Axis Intelligence Research from FDIC data |
| 13 | Illinois | 68 | 670 | 10.1 | $4,495.4 million | 12.17% | Axis Intelligence Research from FDIC data |
Institutions at end-2007 are FDIC-insured commercial banks plus savings institutions. Failures include banks chartered after 2007, so the rate reads as failures per 100 institutions in place before the crisis. Nevada’s FLSR is withheld because Washington Mutual, recorded at zero loss, is most of the state’s failed assets. California includes Nano Banc; Pennsylvania includes Tioga-Franklin.
Nationwide, the 551 failures since 2008 equal 6.5 per 100 of the 8,534 institutions insured at the end of 2007, according to Axis Intelligence Research. Failures were recorded in 42 states and territories. At the other end of the table, Texas had 13 failures against 659 institutions (2.0 per 100), Iowa 3 against 391 (0.8) and Massachusetts 1 against 181 (0.6).
Georgia’s banks were not only more likely to fail; they failed worse. The state’s 32.07% FLSR means the fund lost almost a third of every failed Georgia bank’s balance sheet, and its $10,987.8 million in losses is the largest of any state outside California. The 2026 LaGrange failure, at 31.82%, matched the state’s long-run severity almost exactly.
Sarah Davis: The state map is a map of construction lending in 2006. Georgia, Florida, Arizona and Nevada had the fastest-growing suburbs and the youngest banks financing them, and when lot values collapsed there was no deposit franchise underneath to absorb the hit. Texas, which had already lived through its own bank collapse in the 1980s, barely shows up. Banks that remember a credit cycle tend to underwrite like it.
What Did the 2023 Bank Failures Cost, and Who Paid?
The five failures of 2023 held $532.2 billion in assets, and the three largest, Silicon Valley Bank, Signature Bank and First Republic Bank, held $532.0 billion between them. According to Axis Intelligence Research, those three banks alone held assets equal to 77.2% of all 507 failures in the 2008 to 2014 wave combined.
Silicon Valley Bank’s run was measured in hours, not weeks. The Federal Reserve’s review of its supervision of SVB reports deposit outflows of over $40 billion on March 9, 2023, with management expecting $100 billion more the next day. About 94% of the holding company’s deposits were uninsured at the end of 2022, and the bank had grown from $71 billion to over $211 billion in assets between 2019 and 2021.
The FDIC file records Silicon Valley Bank’s estimated loss at $18.74 billion and First Republic’s at $15.50 billion. Together, according to Axis Intelligence Research, those two losses of $34.24 billion are 32.0% of every dollar of failure losses recorded from 2008 to October 2026.
The FDIC counts these assets two ways. Its failure file uses the last report before closing; the Quarterly Banking Profile’s Table II-C uses final Call Reports and puts the 2023 failures at $552,539 million. We use the failure file throughout, because the FDIC’s loss estimates sit on the same records, and carry both values in the dataset.
Signature Bank is the record that needs a footnote. The FDIC failure file lists its loss as zero, while the FDIC’s March 2023 release on the sale of its branches to Flagstar Bank projected a loss of approximately $2.5 billion. We publish both and adjust nothing silently: the 2023 FLSR is 6.44% as recorded, 6.91% with the 2023 estimate added, and 8.13% with Signature removed.
Who paid for protecting uninsured depositors in 2023?
The banking industry, through a special assessment. Losses from covering uninsured depositors at Silicon Valley Bank and Signature Bank under the systemic risk exception are being recovered through a special assessment on banks with the largest uninsured deposit bases, rather than through regular premiums. The GAO’s report on the FDIC’s final rule describes an original $16.3 billion assessment at 3.36 basis points per quarter on uninsured deposits above the first $5 billion.
The FDIC’s December 2025 interim final rule updated the figures: the loss estimate for protecting uninsured depositors stood at $16.7 billion as of September 30, 2025, six quarterly collections had raised $12.7 billion, a seventh would bring the total to $14.8 billion, and an eighth of about $1.9 billion closes the gap. The assessment falls on 110 banking organizations, down from 114 in the original rule.
Sarah Davis: 2023 rewrote the size table in one quarter. A $209 billion bank lost 8.97 cents per dollar, a normal number for its size, but the bill for the uninsured depositors went to 110 banking organizations through a special assessment instead of staying with the fund. The cost of a run on a bank with 94% uninsured deposits was socialized across the industry’s biggest balance sheets. That precedent is the most expensive thing 2023 produced, and it doesn’t appear in any failure-cost column.
Is Another Wave of Bank Failures Coming? Problem Banks and the Deposit Insurance Fund in 2026
The leading indicators point the other way. The FDIC’s second-quarter 2026 Quarterly Banking Profile counted 47 banks on its Problem Bank List at June 30, 2026, 1.1% of the 4,238 insured institutions and down from 59 a year earlier. Problem banks are those with a CAMELS composite rating of 4 or 5, and the FDIC describes 1 to 2 percent of banks as the normal range outside a crisis. The current count sits below every year-end reading from 2023 to 2025, which climbed from 52 to a high of 66 at the end of 2024 before easing to 60.
| Date | Problem banks | Source |
|---|---|---|
| Dec 31, 2019 | 51 | FDIC QBP, Table II-C |
| Dec 31, 2020 | 56 | FDIC QBP, Table II-C |
| Dec 31, 2021 | 44 | FDIC QBP, Table II-C |
| Dec 31, 2022 | 39 | FDIC QBP, Table II-C |
| Dec 31, 2023 | 52 | FDIC QBP, Table II-C |
| Dec 31, 2024 | 66 | FDIC QBP, Table II-C |
| Dec 31, 2025 | 60 | FDIC QBP, Table II-C |
| Jun 30, 2026 | 47 | FDIC QBP, Table II-C |
The Deposit Insurance Fund rose $3.7 billion in the quarter to $161.1 billion, driven by $2.9 billion of assessment income, and its reserve ratio reached 1.48% of insured deposits. The fund has added $44,179 million since June 2023, when it stood at $116,968 million and a reserve ratio of 1.11% after absorbing the 2023 failures, by Axis Intelligence Research’s calculation on QBP data.
Six failures against 47 problem banks is in line with a system clearing out a handful of weak lenders, not one under strain. The fund’s exposure is a different question. The FDIC estimates $8,693,129 million of uninsured deposits at June 30, 2026, 44.38% of deposit liabilities, against $10,895,042 million insured; the breakdown by bank size is on the Axis US banking statistics page. Uninsured money is what left first in 2023. The standard maximum deposit insurance amount remains $250,000 under 12 CFR 330.1.
The full list of closed institutions, acquirers and dates is kept on the FDIC’s failed bank list. For the operational risks that can push a weak bank over the edge, the Axis financial services cybersecurity statistics and banking fraud statistics pages track incident and fraud losses; failures outside banking are covered on the Axis corporate bankruptcy statistics page. The largest surviving US banks, and how their deposit franchises score, are ranked on the Axis US banking statistics page; the Axis European banking statistics page covers the other side of the Atlantic, the private banking statistics page covers the wealth franchises of the largest groups, and the digital banking statistics page tracks how customers now bank.
Sarah Davis: A falling problem-bank count and a rising fund balance are the quiet signals that matter. The 2026 failures cost the fund 0.15% of its balance, a rounding error for an insurer with $161.1 billion. What would change the picture is not another small Georgia lender. It is a mid-sized bank with a large uninsured deposit base and a securities book underwater, because that is the combination that turned three banks into a $34.24 billion loss in 2023.
What Would Change These Numbers Before Year-End 2026?
Any new failure updates the 2026 table and totals the day the FDIC announces it. The FDIC’s third-quarter Quarterly Banking Profile, expected in late November, will reset the Problem Bank List and the Deposit Insurance Fund figures. Loss estimates for every receivership, including the preliminary figures for Tioga-Franklin and Nano Banc, are revised as the FDIC sells retained assets; we update the FLSR whenever the FDIC failure file changes.
Methodology
Scope. This page covers failures of FDIC-insured banks and savings institutions from 1934 through October 12, 2026. Open-bank assistance transactions are excluded. Industry context figures are as of June 30, 2026.
Sources. Failure records, assets and deposits at failure, estimated losses, resolution types, charter classes and states come from the FDIC BankFind Suite failures API, retrieved October 12, 2026. The two failures not yet in that file come from FDIC press releases dated August 21 and September 25, 2026, with closing details from the Pennsylvania Department of Banking and Securities and the OCC. State institution counts for December 31, 2007 come from the FDIC’s historical bank data (commercial banks plus savings institutions). The 2023 special assessment figures come from the GAO and the Federal Register; SVB run data from the Federal Reserve Board; the deposit insurance limit from the eCFR; the material loss threshold from 12 U.S.C. 1831o. Problem bank, fund, insured and uninsured deposit and industry totals come from the FDIC’s second-quarter 2026 Quarterly Banking Profile (Tables I-A, I-C, II-C and III-C) and its time series spreadsheets, and match the canonical figures on the Axis US banking statistics page. The end-2007 institution count of 8,534 matches in both BankFind and the QBP time series. No figure comes from a ranking publisher, an aggregator or memory.
Formulas. FLSR = estimated loss to the insurance fund ÷ total assets at failure × 100, summed across failures in a group (dollar-weighted). Failures per 100 institutions = failures from 2008 to October 12, 2026 ÷ insured institutions at December 31, 2007 × 100. Average loss per failure = total estimated loss ÷ number of failures. Shares of 2008 to 2026 losses use all recorded failure losses in that period, including the two 2026 press-release estimates.
Scope notes. Estimated losses are the FDIC’s current estimates and change as receiverships wind down; 2026 figures from press releases are preliminary. Twelve failures since 2008 are recorded at zero loss, including Washington Mutual, Signature Bank and Guaranty Bank; we use the file as recorded and show sensitivity cases where it matters. Pre-1980 losses are not recorded in the FDIC file. Three calculations were declined and are published as retracted rows in the dataset: an FLSR before 1980, a Nevada state FLSR, and a per-failure median FLSR.
About This Dataset
The full fact table behind this page, 1,026 rows covering every failure year since 1934, all six 2026 failures, the ten largest failures, FLSR by year, era, size, resolution type, charter class and state, 2007 institution counts by state, the 2023 special assessment, Problem Bank List history, source discrepancies and declined calculations, is available as bank-failures.csv. Each row carries the value, unit, as-of date, source organization, document, URL, retrieval date, a primary-source flag and a method note for every Axis Intelligence Research calculation.
- License: CC BY 4.0. Free reuse with attribution to Axis Intelligence Research.
- Mirrors: Hugging Face, Kaggle and GitHub under the same file name.
- Citation line: Axis Intelligence Research, Bank Failures 2026, 2026.
Cite This Page
APA: Axis Intelligence Research, & Davis, S. (2026, October 12). Bank failures 2026: How many banks failed, every year since 1934, what it cost and the Failure Loss Severity Ratio. Axis Intelligence. https://axis-intelligence.com/bank-failures/
MLA: Axis Intelligence Research, and Sarah Davis. “Bank Failures 2026: How Many Banks Failed, Every Year Since 1934, What It Cost and the Failure Loss Severity Ratio.” Axis Intelligence, 12 Oct. 2026, axis-intelligence.com/bank-failures/.
Chicago: Axis Intelligence Research, and Sarah Davis. “Bank Failures 2026: How Many Banks Failed, Every Year Since 1934, What It Cost and the Failure Loss Severity Ratio.” Axis Intelligence, October 12, 2026. https://axis-intelligence.com/bank-failures/.
Bank Failure Questions, Answered
Is my money safe if my bank fails in 2026?
Deposits are insured up to the standard maximum of $250,000 under 12 CFR 330.1. In 508 of 549 failures in the FDIC file since 2008, a buyer took over both insured and uninsured deposits, and accounts moved to the acquiring bank.
How many banks have failed so far in 2026?
Six, as of October 12, 2026: Metropolitan Capital Bank and Trust, Community Bank and Trust – West Georgia, Kentland Federal Savings and Loan Association, Small Business Bank, Tioga-Franklin Savings Bank and Nano Banc. Together they held $1.45 billion in assets, with an estimated $243.3 million cost to the Deposit Insurance Fund.
What happens to deposits above $250,000 when a bank fails?
It depends on how the FDIC resolves the bank. In a standard purchase and assumption, the buyer takes uninsured deposits too. In an insured-deposits-only deal, used 15 times since 2008 and for the LaGrange, Georgia, failure in May 2026, the buyer takes only insured balances.
What was the largest bank failure in US history?
Washington Mutual Bank, which failed on September 25, 2008 with $307.0 billion in assets. First Republic Bank ($212.6 billion) and Silicon Valley Bank ($209.0 billion), both in 2023, rank second and third.
How many banks failed during the 2008 financial crisis?
From 2008 through 2014, 507 FDIC-insured banks failed, holding $689.2 billion in assets. The single worst year was 2010, with 157 failures.
Who pays for bank failures in the United States?
The Deposit Insurance Fund absorbs failure losses, and it is replenished by assessments on insured banks: assessment income was $2.9 billion in the second quarter of 2026 alone, according to the FDIC. The $16.7 billion cost of protecting uninsured depositors at Silicon Valley Bank and Signature Bank is being recovered from 110 banking organizations through a special assessment.
Why does the FDIC lose more on small bank failures?
Small failed banks have little franchise value for a buyer, so the FDIC recovers less. According to Axis Intelligence Research, failures under $100 million cost 25.69 cents per dollar of assets since 2008, against 6.03 cents above $10 billion.
Which state has had the most bank failures since 2008?
Georgia, with 92 failures from 2008 to October 2026, or 26.1 per 100 of its end-2007 institutions. Puerto Rico has the highest rate, with 4 failures among 10 institutions, according to Axis Intelligence Research.
How many banks are on the FDIC Problem Bank List in 2026?
There were 47 problem banks at June 30, 2026, 1.1% of the 4,238 FDIC-insured institutions, down from 59 a year earlier, according to the FDIC.
What is the Failure Loss Severity Ratio (FLSR)?
The FLSR is an Axis Intelligence Research measure of how much a bank failure costs the Deposit Insurance Fund per dollar of the failed bank’s assets: estimated loss ÷ assets at failure × 100. The 2026 failures carry an FLSR of 16.81%; all failures since 2008, 8.61%.