Field Notes · Vol.04 · U.S. Banking

The family tree of American banking grows inward.

A normal family tree spreads out: one root, many branches. This one runs backward. Twenty-five trunks at the top, every American-owned bank above $100 billion, and two centuries of absorbed names hanging beneath them. Open the tree first. Then read it era by era.

~15 min #banking-history #consolidation #glass-steagall #2026

The thesis · read before the prose

Twenty-five trunks remain. Every branch below them is a bank that no longer exists.

Two ways to read it. In the river, time runs rightward: every bank is a line that begins at its founding and ends at the merger that swallowed it, and the twenty-five lines that reach the present edge are every American-owned banking company above $100 billion. That is the entire list. In the tree view, depth is merger generations instead, down to a water company chartered in 1799 and two banks born in 1784. Every node on the lower levels was once the biggest bank in somebody's town, and every one of them has a story. Search for any name, take the guided tour, or wander: click a name and its story appears below the tree. Some columns are deep with graves; a few, like Goldman Sachs, Northern Trust, and USAA, are nearly bare, and the bare columns are their own kind of story.

drag to pan · scroll or pinch to zoom · click a name for its story
Chapter 1 of 8
Find
Focus
The tree Click any name, or search for one above. Most of them have a story. Or press Guided story and let the tree walk you through two centuries in eight stops.
surviving trunk founding root absorbed name spun back out taken in a crisis yellow tag · resold to another trunk
Oldest living roots · 1784 Massachusetts Bank (inside Bank of America) · 1784 Bank of New York (Hamilton's, still on the door) · 1792 Union Bank, Boston (inside State Street) · 1799 Bank of the Manhattan Co. (inside JPMorgan Chase) · 1812 City Bank of New York (inside Citigroup) · 1825 Commercial Bank of Albany (behind KeyCorp)

Scope: every American-owned banking company above $100 billion in assets, per the year-end 2025 regulatory ranking, with 2026's closed mergers reflected. Foreign-owned U.S. subsidiaries (TD, BMO, HSBC, UBS, Santander and kin) grow from trunks in other forests, and credit unions are a different species; both are out of scope. Depth is merger generations, not calendar distance. Node years mark founding or deal date as noted; sources at the end of this note.

Era I · Merchant Houses · 1799–1932

Two of the oldest roots trace to a single duel.

Aaron Burr chartered the Bank of the Manhattan Company in 1799 by slipping a banking clause into a water-company charter. Five years later he killed Alexander Hamilton, who had founded the Bank of New York and the Treasury. The deepest node on the JPMorgan branch of the tree above begins at one end of a pistol shot. The rest of the century built the merchant houses around it: Drexel and Morgan organizing railroads and steel from Philadelphia and New York, thousands of single-office banks spreading west with the country, no wall yet between taking a deposit and selling a bond.

The founding roots, as they appear on the tree

Massachusetts Bank, 1784, and Providence Bank, 1791: the New England pair hiding inside FleetBoston. The Bank of the Manhattan Company, 1799. The Philadelphia Bank, 1803. City Bank of New York, 1812. Hanover Bank, 1851. Wells Fargo the express company, 1852. U.S. Trust, 1853. First National Bank of Chicago, 1863. Travelers of Hartford, 1864. Drexel, Morgan & Co., 1871. Charles D. Barney, 1873. Chase National, 1877. Wachovia of Winston-Salem, 1879. Washington Mutual of Seattle, 1889. Bank of Italy, 1904. Union National of Charlotte, 1908. Salomon Brothers, 1910. Merrill Lynch, 1914. Bear Stearns, 1923. Dean Witter, 1924. Each is a leaf on the tree above, and each was founded as somebody's future.


One more thing this era lacks: a census. Reliable national counting begins with the FDIC in 1934, which is why the spine chart later in this note starts at the wall and not at the roots. The absence is the point.

Era II · The Wall · 1933–1979

They built a wall, and Morgan split in two.

The Banking Act of 1933, Glass-Steagall, drew a hard line between taking deposits and underwriting securities. J.P. Morgan & Co. chose commercial banking. In 1935 the partners who chose the other side walked down the street and opened Morgan Stanley at 2 Wall Street. On the tree above, this is why one name appears on two trunks. The wall did not just split one house; it fixed the shape of the whole canopy for fifty years.

Banking Act of 1933 · Morgan Stanley opens 1935 · Chase Manhattan forms 1955 · Manufacturers Hanover forms 1961

The plateau

For roughly fifty years the bank count barely moves. The FDIC counts about fourteen thousand commercial banks in 1934 and about the same in the early 1980s. Stability, by design. Mergers in this era build regional houses, not national ones: Bank of Manhattan and Chase National join in 1955, Manufacturers Trust and Hanover Bank in 1961, National City and First National of New York in 1955. The plateau held because the law held it: unit-banking rules kept most banks to a single office, and interstate banking was effectively illegal. A bank could grow old, but it could not grow far.

Era III · The Roll-Up · 1980–1999

The count starts to fall. The deals get names.

Deregulation, the thrift crisis, and the 1994 Riegle-Neal interstate act turn a stable industry into a rolling wave of mergers. The commercial-bank count drops from 14,483 in 1984 to fewer than 9,000 by 1999. Sandy Weill assembles Travelers out of Commercial Credit, Primerica, Shearson, Smith Barney, and Salomon. Chemical swallows Chase and takes its more prestigious name. NationsBank buys Barnett, then BankAmerica itself.

RIEGLE-NEAL 1994  ·  CHEMICAL+CHASE 1996  ·  FIRST UNION+CORESTATES 1998  ·  NORWEST+WELLS 1998  ·  NATIONSBANK+BANKAMERICA 1998  ·  CITICORP+TRAVELERS 1998  ·  GLBA 1999

1998: the year the tree took its modern shape

Three of the five biggest trunks assume their present form in a single year, and a fourth (First Union plus CoreStates, the then-largest bank merger in U.S. history) sets up the branch Wells Fargo will absorb a decade later. The Citicorp-Travelers merger is illegal under Glass-Steagall the day it is signed; the Fed grants a temporary waiver, and Congress repeals the wall in 1999 with Gramm-Leach-Bliley. The bet, sold as the financial supermarket, is that one firm should hold your checking account, your mortgage, your broker, and your insurer. The next era tests the bet.

Era IV · The Crisis · 2000–2009

The piece arrives at its own house style at the moment of maximum stress.

J.P. Morgan Chase forms in 2000. Bank One brings Jamie Dimon in 2004. Then, across roughly six months in 2008, the tree loses more famous names than in any other stretch of its history. Bear Stearns in March. Lehman, Merrill, Washington Mutual, and Wachovia between September and October. Countrywide already gone in January. The last two pure investment banks convert to bank holding companies, ending the separation of 1935.

This is the band where the tree above turns red. Nearly every crisis-colored node dates to 2008 and 2009; the only others, First Republic and Silicon Valley Bank, fell together in the spring of 2023.

Figure · Spine

One curve, assembled from five eras. FDIC-insured commercial banks, 1934–2024.

Values approximate, per FDIC historical banking statistics; series shown is insured commercial banks.


The rescues themselves became legend. Mitsubishi UFJ recapitalized Morgan Stanley with $9 billion delivered as a physical check on Columbus Day 2008, because the wires were closed for the holiday. Washington pressed JPMorgan to consider absorbing Morgan Stanley too; Dimon declined, already digesting two crisis purchases. And the next branch transfer was pre-loaded: the option-ARM book Wachovia bought with Golden West in 2006 detonated on schedule, and the East Coast's fourth-largest bank changed hands in a single week.

Era V · The Concentration · 2010–Now

The trunks rested. Then the buying started again.

After 2008 the canopy went still: the giants digested, new-bank formation fell off a cliff, and the count kept sliding by merger. First Republic went to JPMorgan Chase in 2023, the second-largest failure in history, resolved with the 2008 playbook taken back off the shelf. Then in 2025 the tier below the trunks started consolidating at a pace not seen since the roll-up era, and by 2026 every policy dial that governs this tree points the same direction: inward.

14,483commercial banks · 1984
3,941commercial banks · year-end 2024
$49Bbank M&A announced · 2025
1,200+predecessors inside JPMorgan alone

How banks die now

The failure curve tells the whole modern story in three shapes. The savings-and-loan era killed banks by the hundred: about 157 failed in 2010's echo alone, and the 1980s peak ran higher still. The 2010s wound that down to zero: not one FDIC-insured bank failed in 2018, 2021, or 2022, and the country ran an 800-day streak without a single failure. Then March 2023 broke the streak with a new shape entirely: only five banks failed all year, but they held $549 billion in assets, the worst year in American history measured in dollars. Silicon Valley Bank died in roughly a day at the speed of a group chat. Since then the old pattern resumed at miniature scale: two failures in 2024, two in 2025, two so far in 2026, most of them small banks taken down by suspected fraud rather than by credit. Banks no longer die of disease in large numbers. They either get eaten healthy, or one giant dies suddenly of deposit flight.

Figure · Failure rate

Bank failures per year, 1980–2026. Count collapses; size concentrates.

Approximate FDIC-insured failures per calendar year; 1943–1979 ran single digits most years and is omitted. 2026 is year to date. See the FDIC failed-bank list in sources.

The bailout ledger

The word bailout hides an accounting that is stranger than the politics. TARP disbursed about $245 billion into bank programs in 2008 and 2009; through repayments, dividends, and warrant sales, taxpayers recovered roughly $275.8 billion, a profit of about $30 billion on the bank side specifically. The losses lived elsewhere in TARP, mostly in housing programs that were never designed to be repaid. The 2023 crisis ran on a different chassis entirely: no TARP, no public capital injections. Uninsured depositors at Silicon Valley Bank and Signature were covered under a systemic risk exception, the industry itself was billed for it through an FDIC special assessment, the Fed lent against bonds at par through a new facility, and First Republic was resolved the old-fashioned way, by selling it to the biggest trunk over a weekend. The state has largely stopped writing checks. It arranges marriages instead, and every marriage adds a node to this tree.

2008–09 · TARP disbursed to bank programs$245B
Recovered from bank programs (repayments, dividends, warrants)$275.8B
Peak Fed crisis lending to a single firm (Morgan Stanley, 2008)$107.3B
2023 · public capital injected into failing banks$0
2023 · resolution method of choicesale to a larger bank
2023 · who paid for the deposit backstopthe industry, via FDIC special assessment

TARP figures per U.S. Treasury. 2023 mechanics per FDIC and Federal Reserve resolutions of the March 2023 failures.

2026: every dial set to inward

The current environment is a strange machine when read from the tree's point of view. The Fed has held its target at 3.50 to 3.75 percent for five straight meetings, but the July 2026 vote ran 9 to 3 with the dissenters wanting a hike, and markets price the next move up, not down. The long end is expensive: the 30-year Treasury sits above 5 percent, which pressures bank bond books the same way 2023 began. Meanwhile the regulatory dials all turned the other way. In March 2026 the agencies rescinded the 2023 Basel endgame framework and re-proposed capital rules that cut requirements outright: roughly 4.8 percent relief for the largest banks and 7.8 percent for smaller ones, a full reversal from the 19 percent increase proposed three years earlier. The Justice Department has withdrawn from its bank-specific merger guidelines. Approvals that once took a year now clear in months: every major deal announced in late 2025 had closed by early 2026.

The deals followed the dials. Bank M&A announced in 2023 totaled about $4 billion. In 2024, $16 billion. In 2025, more than 180 deals worth $49 billion, and that excludes the largest of all: Capital One's $35.3 billion absorption of Discover, closed in May 2025, which made Capital One the sixth-largest bank in the country and pulled a former Sears credit card, spun off this very tree in 2007, into a different canopy entirely. Fifth Third bought Comerica and crossed $250 billion in assets. Huntington bought Cadence and did the same. The tier just below the giants is assembling itself into trunk-sized things, with cheaper capital requirements, faster approvals, and a hawkish rate backdrop that punishes the small and the slow. The open questions of 2026 point the same way: deposit insurance reform stalls, stablecoin charters court the same deposits, and commercial real estate stress sits heaviest exactly where the buyers are hunting, in the regional tier.


One trunk, twelve hundred graves

JPMorgan Chase counts more than 1,200 predecessor institutions in its own corporate history. That is the thesis in a single number: one surviving name, twelve hundred absorbed ones. Open any trunk on the tree and the pattern repeats at smaller scale. Bank of America carries a 1784 Boston charter and a 1904 immigrant bank from San Francisco. Wells Fargo carries Philadelphia's first bank. Citigroup carries a charter signed days before the War of 1812. Morgan Stanley, the youngest trunk, already holds a brokerage born inside Sears.


The wall's round trip

Built in 1933. Cut the House of Morgan in two in 1935. Torn down in 1999. By 2008 the last banks it had separated crossed back over, and by 2026 the capital rules written after that crisis are being unwound too. Seventy-three years to complete a circle, and eighteen more to start walking it again. The count is a tenth of what it was, the whole canopy above $100 billion is twenty-five names wide, and five of those names hold nearly two-thirds of it. Nothing in this tree's history suggests it knows another direction to grow.

Sources

  1. FDIC Quarterly Banking Profile · Q4 2024 industry trends: 3,941 insured commercial banks
  2. Federal Reserve History · Gramm-Leach-Bliley (1999); bank count fell from 14,000+ in 1984 to under 9,000 by 1999
  3. Federal Reserve History · Banking Act of 1933, the Glass-Steagall separation
  4. FDIC 2024 Annual Report appendices · failure resolutions and key statistics
  5. Bankrate · List of failed banks 2009–2026, incl. the 800-day streak and 2026 YTD
  6. American Banker · 15 most recent failures, through July 17, 2026
  7. U.S. Treasury · TARP bank programs: $245.1B disbursed, $275.8B recovered
  8. ProPublica · Bailout Tracker, TARP program accounting
  9. CNBC · FOMC July 29, 2026: rate held at 3.50–3.75%, three dissents
  10. Freshfields · Basel III endgame re-proposal, March 19, 2026
  11. Holland & Knight · 2026 capital proposals rescind the 2023 framework
  12. Harvard Law School Forum · Bank M&A 2025: 180+ deals, $49B announced
  13. Capital One 8-K · Discover acquisition approvals, April 2025
  14. eMarketer · Fifth Third–Comerica and the regional consolidation wave
  15. JPMorgan Chase · Our History: 1,200+ predecessor institutions since 1799
  16. Encyclopedia.com · Bank One history; City National of Columbus opened Oct 29, 1929
  17. Morgan Stanley · Our History: founded 1935 as a partnership of 13
  18. Britannica Money · Bank of America corporate history
  19. Britannica Money · Citigroup: City Bank of New York (1812) through the Travelers merger
  20. North Carolina History Project · The Wachovia Corporation, 1866–2008
  21. Federal Reserve · Large commercial banks ranked by assets (the wider canopy's scope)
  22. U.S. Bank · Our history: national charter No. 24, July 13, 1863; the four-way roll-up; Union Bank, 2022
  23. Truist · BB&T and SunTrust complete merger of equals, December 9, 2019
  24. CSMonitor/AP · The Coca-Cola formula leaves the SunTrust vault after 86 years, 2011
  25. NPR · PNC to buy Riggs, the bank of presidents, after the laundering scandal
  26. Goldman Sachs · A brief history: 1869, J. Aron (1981), SLK (2000), bank holding company (2008)
  27. Wikipedia · BNY: Bank of New York (1784), Irving Trust (1988), Mellon merger (2007)
  28. PYMNTS · BNY Mellon rebrands as BNY, 2024
  29. Business Wire · Schwab completes the TD Ameritrade acquisition, fall 2020
  30. FundingUniverse · State Street: Union Bank chartered 1792, signature of John Hancock
  31. Britannica Money · American Express: Wells, Fargo and Butterfield, 1850
  32. Encyclopedia.com · American Express: Shearson (1981–1993), the Lehman spin-off (1994)
  33. MarketsWiki · Capital One: the Signet spin-off (1995), Hibernia, North Fork, ING Direct
  34. Federal Reserve · Order approving Citigroup–Golden State Bancorp (CalFed), October 2002
  35. Fifth Third · Merger with Comerica completed, February 2, 2026; ninth-largest U.S. bank
  36. PR Newswire · Huntington completes the Cadence merger, early 2026
  37. First Citizens · Whole-bank purchase of Silicon Valley Bridge Bank, March 27, 2023
  38. Congressional Research Service · Government assistance for GMAC/Ally: $17.2B TARP, 74% Treasury stake
  39. M&T · People's United acquisition completed, April 2022; Wilmington Trust joined 2011
  40. Encyclopedia of Cleveland History · KeyBank: Society for Savings (1849) and the 1994 Society–KeyCorp merger
  41. Synchrony · Completion of separation from GE, 2015
  42. Citizens Financial Group · Company history: Providence 1828, the RBS era, the 2014 IPO
copper.vision · field notes / 04
read ··· times
David Sierra