Bank failures and supervision, 2008 to 2026: fewer failures, larger shocks, half as many banks

We counted every FDIC-insured bank failure since 2008 and read the FDIC's problem-bank and insurance fund figures to see what they say about bank risk now.

The records behind this report

The FDIC keeps three public records that together describe bank risk over time. The Failed Bank List names every insured bank that has failed since October 1, 2000, with its city, state, acquirer and closing date [1]. The BankFind Suite failures data adds, for each failure, the bank's total assets from its last financial report before closing and the FDIC's estimated loss to the Deposit Insurance Fund (DIF) [2] [3]. The Quarterly Banking Profile (QBP) reports, every quarter, how many institutions are insured, how many sit on the "Problem Bank List" and how large the DIF is relative to insured deposits [7] [8].

Read together, they show two very different kinds of stress since 2008: a long wave of hundreds of small and mid-sized failures after the financial crisis, and a short spring in 2023 when three very large banks failed within weeks. They also show a banking system that is about half the size, by number of institutions, that it was in 2008 [9].

Data and method

Failure counts: we downloaded the Failed Bank List CSV from fdic.gov on November 16, 2026 and counted rows by the year of the "Closing Date" column; it holds 578 rows, the latest being Nano Banc of Irvine, California, closed September 25, 2026 [1]. Assets and estimated loss: we queried the BankFind Suite failures endpoint at api.fdic.gov for failure years 2008 to 2026 and kept only records whose resolution field (RESTYPE) reads "FAILURE"; we excluded the 13 records marked "ASSISTANCE", which are open-bank assistance transactions rather than closings [2]. We summed the QBFASSET field (total assets, thousands of dollars) and the COST field (estimated loss, thousands of dollars) by FAILYR [2] [3].

  • The two counts agree for every year from 2008 to 2025 and for the four 2026 failures through July 17, 2026: 549 failures in BankFind against 549 matching rows in the CSV [1] [2].
  • The BankFind index we queried was built on August 25, 2026 and holds four 2026 failures; the CSV and the FDIC's 2026 summary list six [2] [4]. For the two missing banks we used the figures in the FDIC's closing press releases: Tioga-Franklin Savings Bank, total assets of $68 million and an estimated DIF cost of about $5.5 million [6]; Nano Banc, total assets of $736 million and an estimated DIF cost of about $114 million [5]. These are preliminary estimates that the FDIC says will change [5] [6]. The 2026 row and the total row of the table therefore mix BankFind-recorded figures with these preliminary press-release figures, and their estimated-loss totals are preliminary.
  • FDIC defines estimated loss as the amount disbursed from the DIF to cover insured depositors minus the amount it expects to recover from the receivership [3]. Twelve failures from 2008 to 2026 are recorded as 0 in the FDIC dataset's estimated-loss field, including Washington Mutual Bank (Including its subsidiary Washington Mutual Bank FSB) (2008) and Signature Bank (2023); we report the field as recorded and did not substitute other estimates [1] [2].
  • Institution counts come from the "Number of institutions reporting" row of the QBP time series workbook for the second quarter of 2026; problem-bank counts and DIF figures come from the charts in the FDIC's statement on the same quarter [7] [9].

Amounts are nominal dollars, not adjusted for inflation, and rounded to two decimals in billions. Anyone can reproduce the table by repeating the same CSV download and API query; later FDIC updates to estimated losses will move the cost column.

Failures by year: a crisis wave and a 2023 spike

Bank failures by year of closing, total assets and estimated loss to the Deposit Insurance Fund, in billions of dollars (2026 and total rows include the FDIC's preliminary press-release figures for Tioga-Franklin Savings Bank and Nano Banc)
YearFailuresTotal assets ($bn)Estimated loss ($bn)
200825 [1]371.95 [2]17.73 [2]
2009140 [1]169.71 [2]25.80 [2]
2010157 [1]92.08 [2]15.76 [2]
201192 [1]34.92 [2]6.37 [2]
201251 [1]11.62 [2]2.38 [2]
201324 [1]6.04 [2]1.20 [2]
201418 [1]2.91 [2]0.38 [2]
20158 [1]6.71 [2]0.85 [2]
20165 [1]0.28 [2]0.04 [2]
20178 [1]5.08 [2]1.01 [2]
20180 [1]0 [2]0 [2]
20194 [1]0.21 [2]0.03 [2]
20204 [1]0.45 [2]0.09 [2]
20210 [1]0 [2]0 [2]
20220 [1]0 [2]0 [2]
20235 [1]532.23 [2]34.30 [2]
20242 [1]5.97 [2]0.75 [2]
20252 [1]0.13 [2]0.04 [2]
2026 (to November 16)6 [1]1.45 [2] [5] [6]0.24, preliminary [2] [5] [6]
Total, 2008 to 2026551 [1]1,241.74 [2] [5] [6]106.97, preliminary [2] [5] [6]

The count tells the post-crisis story: 465 of the 551 failures since 2008, or 84%, came in the five years from 2008 to 2012, with a peak of 157 in 2010 [1]. There were no failures at all in 2018, 2021 or 2022 [1]. Since then the count has stayed in single digits: five in 2023, two each in 2024 and 2025, and six so far in 2026, the most in a year since 2017 [1].

Assets tell a different story. The five failures of 2023 held $532.23 billion in total assets, close to the $680.28 billion held by all 465 failures of 2008 to 2012, and 43% of all failed-bank assets since 2008 [2]. Three banks account for almost all of that year: Silicon Valley Bank, closed March 10, 2023 with $209.03 billion in assets; Signature Bank, March 12, 2023, $110.36 billion; and First Republic Bank, May 1, 2023, $212.64 billion [2]. The year's recorded estimated loss of $34.30 billion is the highest of any year in the period [2].

Loss relative to assets moved the other way. Across 2008 to 2012 the recorded estimated loss was 10.0% of failed-bank assets, and in 2010 it was 17.1%; in 2023 it was 6.4% [2]. Part of that gap is how the record is kept: Signature Bank, with $110.36 billion in assets, is recorded as 0 in the FDIC dataset's estimated-loss field [2].

Moraph's view

Moraph's view: the count of failures is a poor measure of risk on its own. A year with five failures was, by assets and recorded loss, the costliest year in this record. Anyone who tracks bank risk by counting closings, whether a treasurer choosing where to hold operating cash or a board reviewing counterparty exposure, would have seen a calm year in 2023 until March.

2026: more failures, all small

The six failures of 2026 range from Kentland Federal Savings and Loan Association in Indiana, with about $3.7 million in assets, to Nano Banc, with about $736.0 million [4]. The others were Metropolitan Capital Bank & Trust, Chicago; Community Bank and Trust - West Georgia, LaGrange; Small Business Bank, Lenexa, Kansas; and Tioga-Franklin Savings Bank, Philadelphia [1] [4]. Each was resolved by an acquiring bank assuming deposits [4]. BankFind records four of them, with $0.64 billion in assets and $0.12 billion in estimated loss [2]. The two newest are not yet in BankFind: the FDIC's closing press releases give Tioga-Franklin Savings Bank $68 million in assets and a preliminary estimated cost of about $5.5 million [6], and Nano Banc $736 million in assets and a preliminary estimated cost of about $114 million [5]. Adding those preliminary figures to the BankFind records gives about $1.45 billion in assets and a preliminary estimated loss of about $0.24 billion for the year [2] [5] [6].

Supervisory statistics: problem banks and fewer banks

Banks on the FDIC "Problem Bank List" (CAMELS composite rating of four or five), selected quarters
QuarterProblem banksShare of banks (%)
2008 Q4 [7]252 [7]3.03 [7]
2011 Q1 (peak) [7]888 [7]11.72 [7]
2019 Q4 [7]51 [7]0.99 [7]
2022 Q4 [7]39 [7]0.83 [7]
2024 Q3 [7]68 [7]1.51 [7]
2025 Q4 [7]60 [7]1.38 [7]
2026 Q2 (latest) [7]47 [7]1.11 [7]

A bank is on the FDIC's Problem Bank List when its CAMELS composite rating is 4 or 5 [7]. The list grew from 252 banks at the end of 2008 to 888 in the first quarter of 2011, 11.72% of all banks, and then fell for most of a decade [7]. It rose again after the 2023 failures, to 68 in the third quarter of 2024, and has since fallen to 47 banks, 1.11% of the total, at June 30, 2026 [7]. The FDIC describes 1% to 2% as the normal range for non-crisis periods [7].

The number of insured institutions fell from 8,304 in 2008 to 4,238 in the second quarter of 2026, a decline of 49% [8] [9]. The fall continues: the workbook shows 4,338 at the end of 2025 and 4,238 two quarters later [9]. In the second quarter of 2026 four banks opened and one failed [7], so most of that quarter's net decline of 41 institutions, from 4,279 at March 31, 2026, came from mergers and other exits rather than failures [9].

The DIF reserve ratio is at its highest level in the FDIC chart series, which starts in 2006 [7]. Its reserve ratio, the fund balance divided by estimated insured deposits, was 1.48% at June 30, 2026, with a balance of $161.1 billion [7]. At the end of 2009 the ratio was negative, at -0.39% [7]; in the first quarter of 2023 it was 1.11% [7].

Moraph's view

Moraph's view: the problem-bank count has been a reasonable early signal for slow credit failures, which is why it peaked a year after failures did. It gave little warning in 2023, when the list held 43 banks in each of the first two quarters while three of the largest failures in the record happened [2] [7]. The spring 2023 failures were liquidity events, and the record shows that a CAMELS-based list should be read alongside deposit concentration and uninsured funding, not instead of them.

How supervision is changing in 2025 and 2026

The agencies that produce these statistics are changing how they examine and rate banks, which will affect how the numbers are read in coming years. As checked on November 16, 2026:

  • On August 27, 2026 the OCC and the FDIC issued a final rule that defines "unsafe or unsound practice" for enforcement and supervision and sets uniform standards for when examiners may issue Matters Requiring Attention, so that material financial risks take priority over policy, process and documentation concerns [11].
  • On April 7, 2026 the OCC and the FDIC issued a final rule that removes reputation risk from their supervisory programs and bars them from criticizing or acting against a bank on that basis [12].
  • On May 19, 2026 the FFIEC asked for comment on proposed revisions to the CAMELS rating system, aimed at material financial risk and greater transparency, while keeping the basic framework; comments were due within 90 days of Federal Register publication [14]. We found no final version as of November 16, 2026.
  • The Federal Reserve finalized changes to its rating framework for large bank holding companies on November 5, 2025, under which a firm with no more than one "deficient-1" component rating counts as "well managed" [16]. Its updated statement of supervisory operating principles, posted on September 24, 2026, supersedes the original statement dated October 29, 2025 and tells examiners to prioritize a firm's material financial risks rather than devote excessive attention to processes, procedures and documentation that do not pose a material risk [15].
  • On January 22, 2026 the FDIC replaced its Supervision Appeals Review Committee with an independent Office of Supervisory Appeals [13].

Two FDIC actions relate directly to the 2023 failures. A staff study released May 14, 2026 used transaction-level data from Silicon Valley Bank, Signature Bank and First Republic Bank and found that depositors with substantial uninsured funds were far more likely to run, while fully insured retail depositors generally did not [17]. On October 30, 2026 the FDIC Board proposed rescinding its part 370 recordkeeping rule for banks with two million or more deposit accounts and moving those banks to the section 360.9 requirements [18].

Moraph's view

Moraph's view: the CAMELS revision matters most for this record. Because the Problem Bank List is defined by CAMELS composite ratings, a change in how ratings are assigned can move the count without any change in bank condition. Readers comparing problem-bank counts across 2027 should check whether the revised framework is in force before treating a rise or fall as a change in risk. The shift toward material financial risk also means fewer findings about governance and documentation, which, in our reading, is where many technology and operational weaknesses have tended to be recorded.

What to watch

The FDIC publishes the QBP around late May, late August, late November and late February [10], so the third-quarter 2026 profile with an updated problem-bank count is due within weeks of this report. As of November 16, 2026 the BankFind failures index does not yet include Tioga-Franklin Savings Bank or Nano Banc; when it does, their recorded figures may differ from the preliminary estimates used here [2] [5] [6]. The CAMELS proposal and the part 370 proposal are both open items [14] [18].

Moraph's view

Moraph's view: the decline in the number of banks has procurement consequences: every merger consolidates core systems, vendor contracts and deposit relationships.

What this means for you

  • Bank treasurers and CFOs: Track uninsured and concentrated deposits as closely as credit quality: the FDIC's study of the 2023 failures found that depositors with large uninsured balances, and the largest depositors in particular, were the ones who ran [17].
  • Bank risk and compliance leaders: Expect fewer examination findings about process and documentation and more about material financial risk under the OCC and FDIC final rule on unsafe or unsound practices and MRAs [11]; the FDIC's new Office of Supervisory Appeals is the route for disputing material supervisory determinations [13].
  • Technology leaders at large deposit banks: Follow the FDIC's October 30, 2026 proposal to rescind part 370 for banks with two million or more deposit accounts [18]; if adopted, it would change what deposit insurance calculation systems those banks must build and maintain.
  • Public agencies, corporates and vendors that hold deposits or rely on bank partners: Read the quarterly FDIC figures for direction, not reassurance: the problem-bank count was 43 in the first quarter of 2023, the quarter in which two of the three largest 2023 failures occurred [2] [7].

Moraph predictions

Every prediction is Moraph's view, not a fact. Each has a date, a confidence level (70% means we expect to be wrong about three times in ten) and a test anyone can apply. We score them on this page when the date comes and never remove one. How we make and score predictions.

  1. P1 By Confidence 80% Open

    Fewer than 10 FDIC-insured banks fail with a closing date in calendar year 2027.

    How we will check: Download the FDIC Failed Bank List CSV in January 2028 and count rows whose Closing Date falls in 2027. Right if the count is 9 or fewer.

    Basis: Failures were 5 in 2023, 2 in 2024, 2 in 2025 and 6 in 2026 through November 16; the last year with 10 or more was 2014 [1]. The Problem Bank List fell to 47 banks at June 30, 2026 [7].

  2. P2 By Confidence 85% Open

    No FDIC-insured bank with total assets of $10 billion or more fails in calendar year 2027.

    How we will check: For every 2027 closing on the FDIC Failed Bank List, read total assets in the FDIC closing press release or the BankFind failures QBFASSET field. Right if none is $10 billion or more.

    Basis: Since 2008, failures of that size occurred in 2008, 2009, 2010 and 2023 only, and none since First Republic Bank on May 1, 2023 [2]; the largest 2026 failure held $736 million [5].

  3. P3 By Confidence 80% Open

    The FDIC's Quarterly Banking Profile for the fourth quarter of 2027 reports 75 or fewer banks on the Problem Bank List.

    How we will check: Read the problem-bank count for the fourth quarter of 2027 in the FDIC's QBP statement or press release, published around late February 2028. Right if 75 or fewer.

    Basis: The list ranged from 39 to 68 banks in every quarter from 2019 through June 2026 and stood at 47 at June 30, 2026 [7]; the QBP is published around late February for the fourth quarter [10]. A revised CAMELS framework, proposed May 19, 2026, could move the count [14].

  4. P4 By Confidence 75% Open

    The FDIC's Quarterly Banking Profile reports fewer than 4,100 insured institutions for the fourth quarter of 2027.

    How we will check: Read the number of insured commercial banks and savings institutions in the QBP press release for the fourth quarter of 2027. Right if below 4,100.

    Basis: The count fell from 4,487 at the end of 2024 to 4,338 at the end of 2025 and 4,238 at June 30, 2026 [9]; four banks opened in the second quarter of 2026 [7].

  5. P5 By Confidence 65% Open

    The FFIEC agencies adopt a final revised CAMELS (Uniform Financial Institutions Rating System) framework.

    How we will check: Look for a final notice revising the Uniform Financial Institutions Rating System from the FFIEC or its member agencies, published in the Federal Register or announced in a press release on or before December 31, 2027. Right if one is issued.

    Basis: The FFIEC requested comment on the revisions on May 19, 2026, with a 90-day comment period after publication [14]; the OCC and FDIC finalized their related rule on unsafe or unsound practices and MRAs within about ten months of an October 2025 proposal [11].

Facts on this page link to their sources. Text marked "Moraph's view", and every prediction, is our own analysis, not a fact or advice. Methodology.

Spotted an error? Tell us and we will check it and correct the piece where needed.

  • FDIC
  • bank failures
  • Deposit Insurance Fund
  • Problem Bank List
  • CAMELS
  • Quarterly Banking Profile
  • bank supervision
  • OCC
  • Federal Reserve