I occasionally deliver an intriguing presentation at bankers conferences called Predictapalooza: It Could Happen. My premise is that at every conference, subject matter experts take center stage to predict what they think will happen in banking. Based on the current fact set, observable past outcomes, and their industry experience.

But what if they get it wrong? Black swan events are difficult to predict and therefore we don’t do it. Events such as, I don’t know, a pandemic. Or a run on a bank initiated by a Tweet. No, we don’t like making those types of predictions because they are so hard to get right. For example, the latest prediction from Jamie Dimon was that unregulated private credit markets could precipitate the next banking crisis. Bank of America’s CEO, Brian Moynihan, said recently that stablecoins could suck as much as $6 trillion out of bank deposits.

The Art of Prediction

Moynihan and Dimon aside, we made our own black swan predictions at a recent banking conference, and I would like to focus on one of them: Bank branches ARE NOT dead.

Ok, that seems less of a black swan as some of our other prophetic predictions. But let’s not forget some industry luminaries that were hammering nails in the branch coffin. Like Ed Crutchfield, former Chairman and CEO of First Union, proclaiming in 1999 that “the branch is dead.”

Many futurists were not as blunt as Crutchfield so they would argue that they meant something different by saying branches would go the way of Blockbuster. And to Crutchfield’s credit, First Union rolled up into what is now Wells Fargo, which has shrunk its branch network from about 5,800 in 2020 to 4,200 today. Futurists love to make bold predictions but fail to mention the end date.

Since branching’s peak in 2009 of nearly 100,000 branches nationwide, banks have collectively culled their networks by 23%, to the current 76,000. With the benefit of the rearview mirror, I may have predicted that the branch is dead… in its current form. Surveys of individual, business, and municipal customers still cite branch location as a top reason for opening an account at a specific bank.

Story continued below…

Interest Rates or Competition

Competition, however, has reduced the spreads that banks can garner from deposit accounts. In the second quarter of 2006, when the Fed Funds Rate stood at 5.25%, the deposit spread plus fees as a percent of deposits was 2.76%. Meaning for every $10 million in deposits, a branch generated $276,000 per year in deposit spread and fees. Note that deposit spreads are calculated using co-terminous funds transfer pricing for a market-based deposit credit, minus actual interest expense paid to customers.

In the second quarter of 2024, when the Fed Funds Rate also stood at 5.25%, the combination of deposit spreads and fees was 2.29% of average deposits. That’s a 47-basis-point, or a 17% decline in revenues afforded the branch.

How did banks respond? Exit branching?

No, they grew the average deposits in them. In 2006, the average branch deposit size was $45 million. In 2024, it was $89 million. The spread plus fees as a percent of deposits was materially down, but the deposits almost doubled. So the 2006 branch generated $1.2 million per year in spread and fees, while the 2024 branch generated $2 million. Banks grew the deposit size of the branch. The above data is from my firm’s database for the hundreds of branches where we measure profitability on an outsourced basis for our clients. There are no top 50 banks in those numbers. They are community banks.

Some of that growth came from consolidation, closing overlapping branches as a result of bank mergers. At branching’s peak in 2009, there were 8,200 banks, twice the number we have today. The largest banks, like Wells Fargo, have been pruning their branch network since that time if not earlier. But the branch being dead? Hardly. Although the largest banks have been net reducing branches, they are still branching and supporting branches in their public statements.

In fact, the direct expense pre-tax profit of the branch remains elevated due to higher interest rates, making deposits more valuable – specifically, reasonably priced deposits. And how do you raise reasonably priced deposits? The branch.

The Risks of Branching

In our “Predictapalooza” presentation, we recommended low to no-cost measures to reduce the risk of branching. One recommendation is to use the “best use of capital” method. Even though branches have very few loans, deposits require capital even though regulators don’t measure it. They still have interest rate risk, liquidity risk, and operational risk. So effective risk management should include capital allocation based on perceived risk of the activity.

For a branch, the capital allocation might be 2% or 3% of deposits, depending on a bank’s risk analysis. Now, does the ROE of the branch exceed another potential investment? If it does, continue to support it. If it doesn’t, can the bank improve its performance? If not, are you willing to pull the plug and invest that capital where it can better serve the bank?

Bank branches ARE NOT dead. The small branch might be dead. The transactional branch is dead. But if you believe that franchise value is delivered, in part, by reasonably priced deposits, how else do you propose getting them?