For decades, FDIC deposit market share data has been one of the banking industry's most important tools for assessing competitive position. It is relied upon by regulators, banks, consultants, and investors to measure market presence and evaluate branch expansion opportunities. However, analysis of two Pennsylvania counties, Lackawanna and Luzerne counties, suggests that the traditional branch-based methodology may be becoming increasingly disconnected from how households actually manage their savings.
These counties offer an interesting case study. Between 2021 and 2026, the number of households increased from approximately 220,300 to 225,000, representing modest but steady growth. At the same time, FDIC-reported deposits attributed to physical bank offices in the two counties increased from roughly $15.8 billion to $16.1 billion. Deposits per household, however, remained essentially flat, moving from approximately $71,870 to $71,490. The compound annual growth rate of deposits per household was actually slightly negative, while household growth exceeded total deposit growth.
The below table indexes 2021 to 100 to demonstrate what I just described.
This raises a fundamental question: what exactly does FDIC deposit market share measure today?
The FDIC Summary of Deposits system attributes deposits to physical banking offices. That methodology was highly logical when households overwhelmingly banked at institutions with local branch networks.
Today, however, consumers routinely place funds with institutions that may have no branches in their county, state, or even region. Online banks such as Ally Bank and SoFi Bank compete directly with community and regional banks for deposits while maintaining little or no branch infrastructure.
Similarly, consumers increasingly hold cash-equivalent assets in brokerage sweep accounts, money market mutual funds, Treasury bills, and other vehicles that sit entirely outside traditional deposit market share calculations.
As a result, a bank's reported deposit market share within a county may tell us where deposits are booked rather than where residents actually store their wealth. A community bank could maintain a stable branch-based market share while losing household wallet share to online banks and investment platforms. Conversely, an online institution could have substantial relationships with local households while appearing to have no market presence at all.
The competitive implications are significant. Regulators use deposit market share extensively in merger and acquisition analysis. Yet if a growing percentage of household savings resides outside locally reported branch deposits, concentrations based solely on branch data may overstate actual market power and understate competition. The market experienced by consumers is increasingly national, not local.
A potential solution would be for banks and credit unions to report deposits based on the ZIP code of the depositor's residence or business location rather than solely the location of the branch where balances are assigned. While implementation would require some data cleanup and privacy safeguards, the resulting information would provide a much more accurate picture of where households and businesses actually place their money. It would also allow banks to make better strategic decisions and regulators to better evaluate competition from digital banks, credit unions, and other federally insured institutions that operate without extensive branch networks.
The Lackawanna-Luzerne data support this theory. Household growth has continued, branch counts have fallen, and deposits per household have not grown. All this while inflation rose a total of 23% since 2021, and money supply (M2) has grown 2.3%, and banks cost of deposits averaged 1.5%-2% annually. Aggregate deposit growth in Lackawanna-Luzerne was less than the interest paid on those deposits.
Those trends are consistent with a banking system in which physical location is no longer the primary determinant of deposit gathering. As banking becomes increasingly digital, a branch-based deposit market share framework may be measuring yesterday's competitive landscape rather than today's reality. The next evolution of market share reporting may require shifting focus from where deposits are booked to where depositors actually live.
Let's make the change.
~ Jeff
zip code reporting is a very good idea
ReplyDeleteI don't think it would be a very big lift.
ReplyDeleteJeff, my old friend, you're both right and very wrong. I think I may have to reach out to you and maybe we have a quick video call to discuss. Zip code is not enough. My single biggest issue with the Summary of Deposits report is that it only recorded half of licensed banking institutions. It was all of the uppercase banks, but none of the lowercase banks. And what are lowercase banks? Credit unions. Their regulators don't require this. Shame on them. But now banking is much more complicated and the Summary of deposit report is an absolute complete failure for anybody who relied on it. In that part, you are very, very right. In simple terms, it needs to have two views, a business view and a consumer view. The business view is all the institutions that hold deposits, and then you need a view of the what I call the consumer view, but that is the institutions that are interfacing with the consumers. And why is that? Banking as a service. But also, what about those deposits that are sitting at, let's say Starbucks for example? They're sitting in whatever bank that has. Well, maybe that should be included as well. In the business view, it sits in the bank where those deposits are, but there should also be a split out of Starbucks because that's where the interface is. So if I got $100, $200, $300, whatever sitting in my Starbucks account, my personal Starbucks account is not that high by the way, that should be reflected. And zip code doesn't go deep enough. In remote areas, yes, but in dense areas, zip code fails. I'm thinking block room census track. Got to get small to that level. Because when you have super look at all the super dense areas across the country, zip code is going to be a complete and absolute failure for some of the use cases that this data is used. And one thing, I saw your comment, I don't think it would be a very big lift. No, not at all. This is something that could have been done decades ago and it wouldn't be a big lift. And with today's tooling, the drop in the bucket. And if any banks ever push back, it's because they're idiots. Because this is something that I, when I was a banker, would have loved to have had. And the fact that credit unions never had to do this, I always felt that was criminal.
ReplyDeletePlease note the preceding comment was a voice transcription using the Google Pixel Rambler feature. I did not go back over and do any type of grammar correction as this was a stream of consciousness comment.
Dave, I think you're right about CUs reporting too. Challenge: FDIC can compel their banks to report this way, they have to get the NCUA to compel CUs to do the same. And there is, of course, going to be slippage with pre-paid Starbucks deposits sitting in Starbucks' bank (as you mentioned) rather than where the customers reside. But let's not let perfect be the enemy of very good. Get the FDIC and NCUA to report by Zip, and you will have much more meaningful information, imo.
ReplyDelete