The nation’s bank branch network lost ground in the third quarter of 2026 with 313 openings and 446 closings, a net loss of 133 branches. That ends a three-quarter run of net gains, the first such run since 2009, which we reported in our second-quarter article. Even though openings were the highest in any quarter since 2020, closings jumped from 247 in the second quarter to 446, the most since the second quarter of 2024.
The past quarter’s losses are smaller than the 175 branches the network lost in the third quarter of 2025, and it was the smallest third-quarter loss in our series, which began in 2010. However, it more than erases the 110 branches gained over the previous three quarters.
The pullback is not evenly spread across the country. Three of the country’s eight regions, the Southeast, the Southwest and the Rocky Mountain states, still added branches over the past five quarters, while the other five lost them. In the third quarter alone, only the Southeast and New England added branches, and the Great Lakes lost more than any other region.
The Growth Streak Ended in the Third Quarter
Table 1 shows quarterly branch openings and closings since 2025’s third quarter. It shows how sharply closings rose after three quarters in which openings outpaced them.
Net gains of 70, 32 and eight branches in the fourth quarter of 2025 and the first two quarters of 2026, respectively, gave way to a net loss of 133 in the third quarter. Openings rose from 255 to 313 in the second quarter, but closings rose faster, from 247 to 446. For the full five-quarter window, the network is down 198 branches, reflecting the sharp losses in the third quarters of 2025 and 2026.
Some Areas of the Country Are Still Seeing Growth While Others Lose Branches
Table 2 shows the net change in bank branches across the country’s eight regions over the past five quarters. It highlights the outsized role the Southeast played in branch growth.
The Southeast gained 166 branches over the five quarters and remains the clear driver of the national picture, with the Southwest (+29) and the Rocky Mountain states (+11) also positive. The Mideast, which includes New York, New Jersey, Pennsylvania, Delaware, Maryland and Washington, DC, lost 116 branches, more than any other region. The Great Lakes and the Far West lost 93 and 84 branches, respectively.
The picture has softened since our second-quarter article. The Southeast’s five-quarter gain is smaller than the 203 branches in our previous analysis, and the Great Lakes and the Plains are losing more now than they were then. The Mideast is the exception, with its five-quarter loss narrowing from 167 to 116.
The third quarter shows how much has shifted. The Southeast added seven branches and New England added two. Every other region lost branches. The Great Lakes lost 55, with 82 closings against only 27 openings, followed by the Plains and the Far West, which lost 31 and 30, respectively. The Southwest and the Rocky Mountain states, which were growing over the five-quarter window, lost 13 and eight branches in the past quarter.
Growth in the Southeast and Southwest Centers on a Handful of States
Table 3 shows bank branch trends in the Southeast and Southwest regions by state, revealing that branch expansion is not spread evenly across the region.
Florida led the way, adding branches in every quarter of the window, followed by Texas, Georgia and North Carolina. Georgia and North Carolina also added branches in every quarter. Texas was the biggest gainer in the Southwest and the only state there with a substantial increase. Alabama and South Carolina also grew. Virginia lost the most branches in the Southeast, and Arizona lost the most in the Southwest. The remaining states in both regions were roughly flat.
The third quarter broke the pattern in several southern states. Tennessee added branches in each of the previous four quarters and lost 13 in the third quarter alone, with 18 closings against 5 openings, which leaves it with a small net loss over the window. Texas lost 5 branches in the quarter after adding 48 over the three before it, and Arizona lost 8. North Carolina (+10), Florida (+8), Alabama (+7) and Georgia (+6) kept adding branches.
Nashville Drops Off the Growth List and New York’s Losses Reflect Scale
Table 4 shows the 10 metro areas with the largest net branch change over the same five quarters. It shows that both gains and losses are concentrated in a small number of metro areas rather than spread broadly across the country.
Four metro areas account for roughly a quarter of net branch losses across all metro areas. New York, Boston, Detroit and Philadelphia lost the most branches over the five quarters, and Detroit and Boston lost the largest share of their networks. Chicago, Los Angeles and Seattle also lost branches, but far fewer. The pattern is strongest in the legacy Northeastern metros, with Detroit as the Midwestern outlier.
The gains are concentrated too. Atlanta led the metro areas that added branches, followed by Dallas-Fort Worth, Charlotte and Miami, with Austin and Orlando close behind. Nashville was among the leaders in our second-quarter article but has since dropped out of the top group. It lost branches in the third quarter, with closings far outpacing openings, and is now down slightly over the five-quarter window.
The New York metro’s branch network is by far the largest in the country, with about 4,060 branches, more than three times the size of Boston’s or Philadelphia’s. Its loss of 53 branches is the largest in the country and equals about 1.3% of its current branch count. The equivalent figures are 1.7% in Philadelphia, 2.3% in Boston and 3.7% in Detroit. New York is a special case because of its size. The metro area spans the five boroughs plus Long Island, the lower Hudson Valley and northern New Jersey, and about two-thirds of its combined five-quarter losses came from outside the five boroughs. Measured against the metro’s full footprint, 53 branches is a small number. The city’s longer view looks different. Manhattan’s branch count fell by nearly 27% between 2012 and 2026 in NCRC’s analysis of FDIC deposit data, a finding The Real Deal cited in a recent story on landlords dividing up former bank and drugstore storefronts.
A handful of large banks account for much of the movement
Branch activity is also concentrated in a small number of banks. JPMorgan Chase opened 266 branches and closed 99 over the five quarters, a net gain of 167. Without Chase, the network would have shrunk by roughly 365 branches over the window instead of 198. Chase opened 52 branches in the third quarter alone, more than any other bank, with the most openings in Florida, North Carolina and Massachusetts.
Several large banks are closing far more branches than they open. Over the five quarters, Wells Fargo closed 80 and opened 18, TD Bank closed 53 and opened none, Bank of America closed 75 and opened 32, and Citizens Bank closed 43 and opened 10.
Fifth Third shows how a merger can move branch counts within a single bank. It completed its acquisition of Comerica on February 1 and has said it would close 75 Michigan branches by September, 55 of them former Comerica locations. Federal Deposit Insurance Corporation (FDIC) filings processed through the end of September record 14 Fifth Third closings in Michigan and 11 in the Detroit metro area, which accounts for most of Detroit’s 13 closings in the quarter. Fifth Third’s openings in the same quarter were concentrated in Georgia, Alabama and Florida. Because the FDIC records filings on its own schedule, more of the announced Michigan closures may be present in fourth-quarter data.
Growth Is Happening for Full-service Branches, Not In-store Locations
Table 5 shows branch openings and closings by service type (full-service brick-and-mortar branches versus in-store/retail locations, such as those inside a supermarket or big-box store) and how they were split between the Southeast and Southwest regions compared to the rest of the country. It shows that growth, even in the fastest-growing markets, happened via brick-and-mortar branches rather than in-store locations, which shrank everywhere.
669 brick-and-mortar branches opened and 465 closed in the Southeast and Southwest, a net gain of 204. In the rest of the country, they opened 643 and closed 973, a net loss of 330. Even in high-growth markets, in-store and retail branch numbers shrank. Nationally, only 30 new in-store or retail branches opened over the five-quarter window compared with 102 that closed, a pattern that held across each quarter regardless of whether the broader branch count was growing or shrinking.
Banking trade press coverage points to a straightforward explanation. In-store branches were built to provide transactional convenience for customers who were already at the store for something else. Online and mobile banking have eliminated most of the reasons for that kind of interaction since routine transactions no longer require showing up in person. When customers do still need to visit a branch, recent reporting on bank branch strategy suggests they prefer a full-service location better suited to private conversations about banking needs rather than a retail store built for foot traffic.
The Regional Split Lines Up With Where Money and Higher Earners Are Moving
This data cannot identify the business decisions behind a bank’s branch strategy. However, the geography trends we’ve pinpointed line up closely with outside migration research. Demographic analysis published by Brookings has tracked domestic out-migration from large coastal metro areas, including New York, Los Angeles, San Francisco, Boston and Seattle, alongside sustained in-migration to Sun Belt metro areas, such as Dallas, Tampa and Raleigh, since the early part of the pandemic. The same research found that much of what looked like a population rebound in big metro areas in 2023 and 2024 was driven mainly by a rise in immigration rather than a reversal of domestic migration rates. The newest available analysis on this migration found that metro population growth is declining again as immigration slows.
IRS migration data, which tracks income rather than head counts, shows the steadier pattern of growth behind the population volatility. The most recent filing-year data available covering 2022 to 2023 shows Florida gaining more than $20 billion in net adjusted gross income from interstate migration and Texas gaining more than $5 billion, while California, New York, Illinois, Massachusetts and New Jersey experienced the largest net losses. This state-by-state pattern matches our branch findings closely; the same handful of states gaining income are the ones gaining bank branches. New York, despite an overall population that has recovered in recent years due to immigration, continues to lose more income to domestic migration than nearly any other state, which is consistent with its outsized share of branch closures.
A note on data revisions
The figures in this article reflect the FDIC’s BankFind Suite structural change data as of the article’s publication date. Banks sometimes file branch opening, closing and relocation records after the quarter has ended. The FDIC assigns these filings a processing date that can fall months after the effective date of the change. As a result, figures for a given quarter can shift in later reports as additional filings are processed. This article reflects the most current data available and updates the prior quarters’ figures accordingly. The five-quarter window shown was refreshed against the newest available data pull and includes revisions to each of the four prior quarters. The net change for the first quarter of 2026 is now a gain of 32 branches, down from 37, and the second quarter is now a gain of 8, down from 13. These revisions to prior quarter counts are a part of normal data processing and are not corrections to the overall methodology.
The table below shows each quarter’s net change as first published, as reported in our second quarter article and as it stands in this edition.
Methodology note. Figures cover FDIC-insured institutions’ full-service branch locations (brick-and-mortar and staffed retail/drive-through) based on FDIC’s OSCR structural change filings for the quarter shown. Regional definitions follow the U.S. Bureau of Economic Analysis’s eight-region scheme rather than the Census Bureau’s four-region scheme. Metro areas are Census Bureau core-based statistical areas assigned by county. Puerto Rico and other US territories (3 openings and 7 closings over the five quarters) are included in national and state totals but not in any region.
Jason Richardson is the Senior Director of Research with NCRC’s Research team.
Photo: © William – stock.adobe.com
