187 U.S. counties have exactly one bank

The count

The FDIC records every insured bank branch in the country and the deposits it holds. Group those branches by the organization that owns them and 187 of the 3,104 U.S. counties and county equivalents come back with a single name. Every branch in the county belongs to one bank holding company. The local deposit HHI in each of them is 10,000, the maximum the index allows.

Another 20 counties have more than one bank on paper but one bank holding at least 90 percent of the deposits, which brings the number of counties where a single organization holds 90 percent or more to 207. Counties with two or fewer organizations number 524. In 2015, the first year in our panel, 143 counties had one bank. The number has grown by 44 in a decade.

Where they are

The largest one-bank county by deposits is Caroline, VA, where Atlantic Union Bankshares Corporation holds $1.1B across 2 branches. The table lists the twenty-five with the most money on deposit; each links to the county page with its full history.

The states with the most

Texas (22), Montana (17), and Georgia (14) lead the list. These are mostly rural counties, which is exactly why the pattern matters: a branch closure or a merger in a county with two banks turns it into a county with one, and the nearest alternative can be an hour away.

Near-monopolies with a second name on the sign

The one-bank list undercounts the problem. A county where a second bank keeps a single branch with a few percent of deposits is competitive in the FDIC's ledger and not in anyone's experience. These are the largest counties where the leader holds 90 percent or more and at least one other organization is present.

What should happen

Bank merger review in the United States is conducted against statewide and metropolitan market definitions, and a denied application is rare enough to be news. The county screen shows where that framework fails: places where the merging parties are the only two banks a resident can drive to. We think a merger that creates or entrenches a one-bank county should be treated as presumptively harmful, with the burden on the parties to show why it is not.

We also think the FDIC should publish this screen itself, annually, with the same grouping to holding companies that we use. The data exists. The agency already collects it. Making the count official would let branch-closure notices, Community Reinvestment Act reviews, and merger applications be judged against a public benchmark instead of a private one.