Jeff For Banks

What is the better banking strategy: low expense v low deposit costs?

Here is a question that dogs us: should we follow a strategy that drives top quartile performance in operating expenses or cost of deposits? If your answer is “yes”, your execution better be flawless. Because the two don’t often go together. In my search, described below, only one bank made top quartile performance in both categories.
The low operating expense bank typically comes with a limited branch network. Bank futurists think this a good thing since branches are millstones around our collective necks. So in order to attract deposits, these banks tend to use premium rates to get people comfortable with not having a nearby branch. 
Low expense banks tend not to have sophisticated commercial banking operations too. Preferring instead to focus on residential real estate lending combined with transactional commercial real estate lending. By transactional, I mean the bank may resort to price to get deals, and not extensive customer relationships, as many of these customers don’t value relationships anyway.
Conversely, low cost of deposit banks tend to have more expansive branch networks to get core retail and business account balances. Anecdotally, the most often heard reason a business customer objects to opening an operating account at a bank is because it lacks a nearby branch. But, admittedly, that could just be an objection that is not the “real reason”.
Also, low cost of deposit banks tend to be heavy commercial lenders, both commercial real estate and business loans. That requires a good product set, and lots of resources. Asset-based lending requires much greater borrower interaction than plain vanilla commercial real estate lending.
So which strategy results in better performance? I went to the numbers (see table), and it’s tough to tell.
I searched for banks of a certain size ($1B – $20B in assets) to eliminate trading anomalies from very large and very small banks. I also controlled for capital levels and asset quality to limit the number of factors impacting pricing. 
The picture is not totally clear on which strategy delivers the better returns or higher trading multiples. To be sure, both sets of top quartile banks are well run and rewarded with relatively high trading multiples. And they have delivered very good results, as indicated by their 3-year price changes.
So, which one is better? I think it depends on each financial institution’s individual circumstances. Do you operate in slow to no growth markets? Perhaps a low operating expense/cash cow strategy is appropriate. 
Is your FI in business-rich markets and your niche is concierge-like service to tech firms? Well the low cost deposit strategy with higher expense ratios may be appropriate. 
Whatever strategy you choose, I implore you… to choose. The zigging and zagging, the all things to all people, the yes, yes, and yes strategies are misallocating resources and driving our financial institutions into homogenized and commoditized, irrelevant, soon to be sold and quickly forgotten memories that our 7,000 brethren that have been sold since the early ’90’s have become.
Are we going to let it happen?
~ Jeff