Part 1: Are Branches Really Becoming Obsolete?

New Call-to-action

Are Branches Really Becoming Obsolete?

The financial services sector has been altered in many ways since the Financial crisis. One of the most interesting, but also most under-studied phenomena, is the reduction in the number of physical bank and credit union branches across the United States over the past 8 years. U.S. banks closed 4,281 branches between 2009 and 2014, a reduction of almost 5%. Similarly, credit union’s closed almost 1,000 branches between 2011 and 2016, a 4% reduction in the number of physical branches.   

Advances in mobile banking technologies have likely reduced the need for physical branches. Increased regulatory costs might play a role as well.  While the causes of branch closings are likely many and varied, and difficult to precisely pin down, in this blog we briefly touch on potential consequences of declining branch locations.  One potential consequence is rooted in the fact that loan markets are surprisingly local, suggesting that the location of branches still matters, even in a technologically advancing world.

To get a better handle on just how local loan markets can be, consider some of the following facts calculated from a sample of direct auto loans. In a sample of 400,000 direct auto loans from 103 institutions around the country, we use a combination of intensive computing power and state-of-the-art geocoding techniques to calculate the actual driving time from a borrower’s address to the location of the branch where the borrower originated a direct auto loan.  The calculation of actual driving times relies on posted speed limits along optimal driving routes and abstracts from driving conditions (i.e. we are unable to adjust for increased driving times if the drive is made during rush hour).

Are Branches Really Becoming Obsolete_image.png

Our calculations suggest that the median borrower lives within a 15.5-minute drive of the physical branch where they originated a direct auto loan.  Though the median drive is 15 minutes, driving distances show some interesting variation. On the low end, 25% of the borrowers in our sample live within 7 minutes of the institution where they originated the loan. Over 75% of the borrowers in our data live within a 30-minute drive of their lending institution.  As a means of benchmarking these driving times, we compare them to average commute times reported in the most recent U.S. census.  Census data indicate that the median commuter drives 24 minutes to work. Only 25% of surveyed workers live within a 20-minute drive while 75% of all surveyed workers living within a 28-minute drive.

Why do any of these driving time calculations matter?  They suggest, importantly, that loan markets are much more local than labor markets. This evidence is also slightly at odds with the narrative that technology is rendering moot the importance of physical branch locations.  Though data on internet-based lending is currently unavailable to us at VE, it is unlikely that the “best” loan made available to borrowers through an internet search is from an institution closer to their house than their place of employment.  Although borrowers likely search digitally to understand the potential menu of available interest rates, at least in the case of direct borrowers, when customers originate a loan, they appear to do so at a branch that is close to home.  Ultimately, these results suggest that search costs play an important role in credit markets.  

Summing up, branch locations are on the decline, but lending is local. If declining branch locations have the effect of increasing search costs for borrowers, this could have at least two effects. First, increased search costs might make borrowers less likely to originate a loan in the first place. Second, increased search costs might result in borrowers accepting loans on terms that are not the best available because declining branch locations could render the cost of searching out the full distribution of available loan offers too high.  

What are the policy prescriptions? Branches can be expensive, and are declining for a reason. But CUs would do well to remember that, when it comes to originating loans, the data indicate that relationships and familiarity seem to matter.  


Taylor Nadauld

Chief Economist