Closing the Branch Costs More Than the Branch

The Obvious Case for Closing

A branch costs money. Premises staff security technology maintenance all of it. Costs are visible tied to a location and easy to compare with the transactions that occur within it

Those transactions have collapsed. You deposit a check by photographing it. You check the balance on the phone in your pocket. Transfers clear instantly. Cash has been declining as a share of purchases for years. Branch visits per customer have fallen sharply over two decades not gradually

Based on transaction cost the conclusion writes itself and thousands of branches have closed using exactly that logic. Compare what a location makes to what it costs to keep the doors open and closing it seems like the obvious decision

What the Transaction Count Misses

The complication is that a branch does two things that the transaction count never captures and both appear on the balance sheet instead of the income statement that most people look at

Collection of deposits. Decades of research on deposit behavior keep finding the same pattern: People open accounts close to where they live and work and the presence of branches closely tracks a bank's local deposit market share. Deposits are the cheapest financing a bank has access to and the gap between what it pays depositors and what they would pay if they borrowed the same money elsewhere is a big part of why banking is profitable

A branch that barely processes transactions may still have a large deposit base that partly disappears if it closes. The real cost of closing it is the cost of replacing that financing and that cost does not appear in the branch's own income statement. It appears in the numbers on the Treasury table months later disconnected from the decision that caused it

Banking for small businesses. Business customers use branches much more than consumers primarily for cash deposits and tend to value having a real person to call. Small business loans are relationship-based and the relationship is typically tied to a location and a name not an app

Deposit Beta and Why Sticky Money Is Worth More

Here's a term that does a lot of the work in this article: deposit beta. It measures how much a bank's deposit rate changes when a benchmark rate say the Federal Reserve funds rate moves by one point. A beta close to 100 percent means the bank has to match the market movement roughly dollar for dollar. A beta of 20 percent means its deposit rate only rises by about a fifth

Retail deposits especially checking accounts and small business operating balances tend to have a low beta. People don't buy the rate on their checking account the same way they buy the rate on a CD. Wholesale financing intermediated deposits and anything explicitly brought together with a rate announcement behave in the opposite way. Their beta runs close to one sometimes above one if a bank is struggling to get financing in a hurry

Let's say the federal funds rate rises a full percentage point over a year just as an illustration. A retail deposit base with a beta of 20 percent sees its cost rise by about 0.2 points.900,000 dollars. The same 100 million dollars the same rate change a very different bill simply because of the origin of the money

That's why sticky retail deposits are worth more than a simple rate comparison suggests. They're cheap today and stay relatively cheap throughout the rate cycle which is exactly when a bank most needs cheap financing. A branch is one of the main places where stickiness is built in the first place

The Formats That Replaced It

Most banks haven't chosen between a full branch or nothing at all. Reformatting has been the real answer

Networks have moved to fewer and smaller locations with different staff more geared toward advisors than tellers. Machines handle routine deposits and withdrawals and staff spend their time on loans advice and trading accounts

Average branch size has decreased substantially and the former specialized teller role has given way to universal bankers who handle a broader range of jobs from a single desk

The economics of that smaller format differ enough that comparing the number of branches over time without adjusting for format change will mislead you. A bank that has half as many branches as it did a decade ago isn't necessarily committed to branch banking. It could occupy roughly the same space in a fraction of the square footage

A Worked Example: Valuing the Deposit Base

Let me specify the compensation with round and clearly illustrative numbers. None of this is data from a real bank. It is a model of the mechanism built so that you can check each step yourself

Start with the branch's own numbers the version that marks it for closure first. Let's say it costs $1.8 million a year to operate: rent staff security technology all included.Let's assume that fees and transaction revenue credited to that location are $1.2 million per year. On net the branch theoretically loses $600,000 per year since $1.8 million minus 1.2 million is $600,000. That loss is the entire reason for closing it and fits on one line on a spreadsheet

Now bring up the deposit base. Let's say this branch has an average of $150 million in retail deposits spread across checking savings and small business accounts. The bank pays depositors an average of 0.4 percent on that money a blended rate between interest-free checking accounts and modestly yielding savings. If the bank were to raise that same $150 million on the wholesale market through something like an FHLB advance or a CDnegotiated let's call that rate 4.4 percent

The difference between those two figures 4.4 percent minus 0.4 percent is 4.0 percentage points. That differential is approximately what a dollar of retail deposits is worth to the bank each year relative to purchasing the same dollar in the wholesale market

The branch doesn't need to keep the $150 million for that value to matter because closing it won't cost the bank the $150 million. Some of it moves to the nearest branch or is simply serviced through the app and stays there. What matters is the runoff: the portion that leaves the bank entirely because it was tied to that specific location that specific relationship manager that specific parking lot

Let's say the settlement after closing is 40 percent toward the high end of what would be expected from a branch with a real portfolio of small businesses and some long-time customers but it's not an extreme assumption. 40 percent of $150 million is $60 million coming directly from the bank

That $60 million now has to be replaced and it is replaced at the wholesale rate instead of the deposit rate. The incremental cost is $60 million times the spread of 4.0 points which equals $2.4 million per year

Put the two effects together. Closing the branch appears to save $600,000 a year based on the branch's own figures. It also costs $2.4 million a year in higher financing costs once the deposits that were lost are replaced. Net effect: 600,000 minus 2.4 million is negative 1.8 million. Closing this branch does not save the bank any money. It costs the bank $1.8 million a year ifThey follow deposits instead of stopping at the branch's own income statement

One more cut is worth making here: What settlement rate would make the closure truly profitable? Set the increased financing cost equal to the apparent savings of $600,000 and solve for the amount in second-hand dollars. 600,000 divided by 4.0 percent is 15 million. 15 million is 10 percent of the base of $150 million. That is the true threshold.With more than a dollar in ten of your deposits leaving the bank for good rather than simply moving to another channel the cost of funding wipes out all the apparent savings and then some. Ten percent is not a high bar to clear which goes a long way toward explaining why closure decisions continue to be reviewed

Case Study: JPMorgan's Branch Bet

The clearest real-world example of a bank acting on this logic rather than simply modeling it on paper is JPMorgan Chase's branch expansion. In 2018 the bank announced that it would enter markets across the country where it had never operated a retail branch a surprising move at a time when the conventional wisdom in banking was that branches represented a declining legacy cost. Over the next few years JPMorgan opened branches in states where it had no prior retail presence advancingtowards something close to national coverage

The logic laid out closely follows the mechanics of this article. Management including Jamie Dimon in letters to shareholders and on earnings calls has repeatedly pointed to branches as the primary driver of deposit growth and a way to gain a customer's entire relationship - checking savings credit cards mortgages - rather than a single rate-bought product.research

I have to be honest about what this case study demonstrates and what it doesn't demonstrate. It shows that a sophisticated well-capitalized bank looked at the branch closure trend sweeping the industry and made the opposite bet in many markets and continued to do so for years which is a real sign of how that bank's own economics work. It doesn't prove that all branches are worth preserving. JPMorgan also closed and consolidated locations where the math was going the other way in the same years it was opening new ones elsewhere. Thelesson is asymmetry not a general rule: the same company simultaneously closed weak branches and opened hundreds of new ones because the decision is genuinely local there is no single verdict on branches as a format

The Regulatory Layer and Where Closures Concentrate

Closing decisions are not purely commercial. Federal law requires banks to give advance notice to regulators and customers before closing a branch and closures in low- and moderate-income areas receive additional scrutiny under community reinvestment obligations

The concern that regulators handle is what is called a banking desertResearch on this has found that branch closures in already underserved areas are associated with reduced lending to nearby small businesses and increased use of higher-cost alternatives the kind of check cashers and short-term lenders that are expensive precisely because they're filling a gap that a branch used to fill

On the community side the pattern is consistent enough to state clearly: Closings are more concentrated in low-income and historically underserved neighborhoods than across a bank's entire network as a whole. That's a description of where closures are landing not a statement about anyone's motive and I'm not going to comment here on intent. It's one of the reasons regulators created a review process specifically around this issue rather than leaving the decision solely up to profitability math.at the branch level

That evidence is exactly why closure decisions in those areas carry regulatory and reputational costs and it's a real part of why banks keep locations open that don't pass an internal profitability test. The regulatory cost is one more number found outside the branch's income statement right next to the funding cost in the example above

The Digital Only Comparison

Branchless digital rivals have demonstrated both the promise and the limit of the alternative model

They acquire customers cheaply operate at lower costs and have grown rapidly. They have also found that raising deposits is more difficult than their early presentations suggested. Without a physical presence deposits are attracted primarily by the rate and the rate is exactly the wrong tool if what you want is the fixed low-beta financing described above. The money that comes in for the rate tends to leave the moment a better rate appears elsewhere

Several digital banks have responded by pursuing primary account status: direct deposit of a customer's paycheck everyday spending small business banking all with the goal of manufacturing the rigidity that structurally builds a branch network without a branch to build it around

This is some of the most useful evidence available about what branches actually do. Companies built from scratch without them have spent years and real money trying to replicate the exact ownership that a branch confers almost as a side effect

Where This Breaks: The Branchless Counterexamples

I want to retract my own argument here because the honest counterexample is real not a symbolic gesture of balance

Some digital-only banks have created really large fixed deposit franchises without branches. Ally Bank has operated a branchless deposit operation for almost two decades and turned it into a large long-lasting foundation not just a pile of hot money chasing the highest rate on a comparison site. Discover funds a large portion of its card and loan portfolio in the same way. Marcus the consumer deposit arm of Goldman Sachs raised tens of billions of dollars without a single branch earmarked primarily for balances.savings instead of checking accounts

That really complicates the neat story of this article. If a branch were strictly necessary for fixed retail deposits these franchises should not exist at their current size and they clearly do. What I think is actually happening is that a branch is one route to rigidity not the only one. A strong brand a genuinely good product real switching costs once someone's history and tax documents are in your application and enough time in the market can generate some of the same rigidity that a branch generates through habit and geography. It seemsthat both paths can converge to a similar outcome through a different mechanism: brand and product loyalty rather than local presence

Where I think the branch still wins and this is a real hedge on my part rather than a certainty is specifically in small business checking and operating accounts the balances that arrive with a paycheck a payroll run a cash deposit that has to physically go somewhere. Savings balances that chase a published rate are exactly the segment that digital banks have shown they can pull together well. The deposits most tied to a physical presence and the ones the example worked on above leans on most are the ones that aretied to doing business locally. That distinction matters more than a general statement that branches are or are not necessary

How I Actually Use This

The way I use this when reading a bank's records is to stop relying on changes in branch count as a headline number alone. A bank that closes 200 branches is not automatically becoming more efficient and a bank that opens 200 is not automatically spending too much. I want to see it compared to deposit growth and the cost of deposits in the same statement not treated as a story in itself

My read is that the most useful line in any bank's disclosures for this specific question is the cost of deposits relative to its peers tracked for a few years alongside the branch count. If the cost of deposits increases faster than its peers as the number of branches decreases that's the trade-off from the worked example above shown in real numbers and it's not a good sign

I admit I got this backwards the first time I saw a bank aggressively cutting branches as its stock got cheaper. I read the closures as pure cost discipline and filed them away as a point in the bull case. In reality it took running numbers close to the above to notice the deposit cost line moving against them at the same time which is what tells you branch counts alone will never get you

I don't think this framework tells you to buy or sell anything on its own and I'm not going to pretend that it does. What it offers is a way to read a branch closing announcement as a financing decision disguised as a real estate decision which is closer to what it really is

What an Analyst Should Look At

A few numbers do most of the work here. Deposits per branch and their trend tell you whether a network is rationalizing well or simply shrinking. The cost of deposits relative to peers tells you whether the loss of branch presence is already being reflected in more expensive financing. The match between consumer and business deposits is also important since business relationships depend more on branches than on consumers

A bank that reduces the number of branches while deposit costs rise faster than its peers is not handling this trade-off well. It is describing it poorly in public one earnings call at a time

The Bottom Line

Branch networks seem obsolete measured by the number of transactions and seem valuable measured by the cost of financing and business relationships and that gap is exactly why closure programs continue to be paused reformatted and revised rather than simply ended

The cost of keeping a branch open appears on one line is easy to find and easy to cut. The benefit manifests itself elsewhere: in the price a bank pays for money it didn't have to borrow. Run the numbers as the example above does and a branch that looks like a $600,000 annual loss on its own income statement may be worth $1.8 million a year to stay open once it tracks deposits instead of stopping at the branch door

JPMorgan bet real capital on that logic by expanding into new markets rather than contracting and digital-only banks like Ally and Discover complicate it by showing that a branch isn't strictly required to collect fixed deposits either. Both are true at the same time. My view is that a branch still does most of the work exactly for small business deposits checking accounts payroll and cash which are harder to gather with a fee announcement and an app alone. That's the part about theledger closing decisions that continue to be quietly rediscovered