When I think about my childhood, the first and most vivid memory I have is of the summer before I turned 12. Although I didn’t realize it at the time, that summer was as close as I would ever come to realizing my dream of becoming a major league baseball player. That is because it was truly my only occupation, or obsession, for those three months. Each day, I would walk out of the house around 8:00 AM heading toward my friend Bill’s back yard, where I would play baseball with a “rag” ball (thank goodness considering how many houses and windows we hit) and my friends Chip, Clint, Jeff, Michael, and whomever else wandered up that day until the sun set and we were forced to go home. My specific job in the league, besides playing, was to serve as its official scorekeeper and statistician, which allowed me to combine my first love, baseball, with my second, numbers, in a way that still makes baseball appealing to me today. As each batter would come to the plate, I was duty bound to announce their season batting average as well as other pertinent statistics, such as how many homeruns they had hit off of the roof of the house next door. This love of baseball and numbers is still the reason I have to stay up until midnight watching “Moneyball” anytime I find it on cable at 10:30 P.M. It is also a trait I have passed along to my progeny, who keep the floors of our home covered with hundreds of baseball cards that they have sorted, studied, and stacked for hours like they had a final on Manny Machado’s slugging percentage the next day. Yes, my wife is an exceptional woman.
Because I love numbers, I subscribe to the philosophy that they don’t lie. However, I’m sure we are all familiar with Twain’s adopted view of statisticians, placing them on par with liars and the damned. Considering my love of numbers and my chosen profession, I’m not exactly comfortable about what such theories may imply about my character and my eternal prospects, Nevertheless, it would not surprise Twain to find out that the Federal Deposit Insurance Corporation (“FDIC”), after conducting an extensive study regarding issues facing community banks, reached a conclusion about the profitability of community banks relative to non-community banks that is contrary to many of the conclusions I have reached in recent blog articles.
To defend myself, before I discuss the FDIC’s conclusions, I think it is important to note a couple of important distinctions between the FDIC’s study and the research I have relied upon. First of all, that study adopts a completely different definition of community banks than the one I have relied upon. For the purposes of its study, the FDIC developed a new research definition of a community bank that was partially tied to asset size (i.e., an indexed maximum asset value that began at $250 million in 1984 and increased to $1 billion in 2010), but also considered “criteria related to traditional lending and deposit gathering activities and limited geographic scope” (e.g., loans to assets > 33%, core deposits to assets > 50%, numbers of offices, and numbers of offices in MSAs and other states). My definition of a community bank for these articles, particularly my first addressing their competitive profitability, has been much more fluid and has been based solely on asset size.
Secondly, the FDIC’s study period of community bank profitability (i.e., 1985 – 2011), in addition to being primarily before the time period I considered (i.e., 2010 – 2014), was also more than five times longer. Therefore, it encapsulated many more market cycles than the data I studied. For that reason, some may argue that it is more reliable because it is more comprehensive; however, I think the counterargument would be that the last five years have been the most relevant for community banks to consider in this new world since they reflect life in the post-apocalyptic world following Dodd-Frank and the Great Recession.
Nevertheless, considering these differences, I thought it was important to address the following conclusions reached by Chapter 4[1] of the FDIC study, some of which were contrary to my analysis:
- A comparison of pretax ROA[2] reveals that non-community banks (i.e., 1.31 % average pretax ROA) outperformed community banks (i.e., 1.02% average pretax ROA) during most of the FDIC study period;
- Non-community banks had greater success in generating noninterest income from a variety of sources (i.e., average of 2.05 % of average assets vs. 0.8 % for community banks over that same period), explaining much of the gap in earnings;
- Community banks, because of their heavy dependence on lending as a source of income and the long term trend toward lower net interest margins, also experienced a significant erosion in its traditional net interest income advantage over the last few years, which also contributed to the gap in pretax ROA; and
- Even though community banks have traditionally been less efficient than non-community banks, this gap, as measured by the efficiency ratio, widened over the FDIC study period. While the gap was only 3.5% between 1985 and 1998, it ballooned to 9.2 % between 1999 and 2011. This was driven by community bank’s decreasing competitiveness in generating revenue along with their decreasing advantage from lower noninterest expenses, which is now almost non-existent despite a long-term noninterest expense advantage of 22 basis points.
That being said, there were also several points made in Chapter 4 of the Community Banking Study that supported conclusions I reached in my analysis. For example, for both community banks and non-community banks, banks headquartered in metropolitan areas had lower pretax ROAs than banks headquartered in nonmetropolitan areas. Community banks headquartered in nonmetropolitan areas averaged a pretax ROA of 1.25% compared with 0.94% for their urban counterparts, while non-community banks headquartered in rural areas averaged a very impressive pretax ROA of 1.88% compared to 1.30% for community banks headquartered in metropolitan areas. Therefore, the “city bank -country bank” dichotomy identified in my previous articles held up in the FDIC study as well. The FDIC study also noted that community banks have almost always incurred lower credit losses than non-community banks, which helped to narrow the overall earnings gap during the latter years of the study. Finally, with both sets of data, there is little argument that larger banks experienced a significant advantage with regards to net overhead and efficiency ratios.
Nonetheless, even the FDIC has concluded that that “while the results show that community banks may benefit from economies of scale, there is no indication of any significant benefit beyond $500 million in asset size, and much of the benefits from scale appear to be achieved for [community banks] as small as $100 million.” [3] Instead, the study reaching this conclusion found that community bank’s decreasing competitiveness with regards to efficiency ratios is more closely related to the growing erosion of their net interest income advantage and their inability to increase their assets managed relative to numbers of employees the same way that non-community banks have. So, once again, maybe size isn’t all that important, at least when you are talking about bank assets. However, the limited size of my buddy Bill’s yard was extremely important since it allowed me to entertain a dream of hitting similar grand slams in major league parks one day, at least for a summer.
[1] FDIC Community Banking Study (December 2012), Chapter 4, “Comparative Financial Performance: Community versus Non-community Banks.”
[2] The FDIC Community Banking Study focused on pretax ROA as opposed to ROA after tax, which was the basis of my conclusions in my first blog article. They stated that such a focus better facilitated comparisons between banks organized as C corporations (i.e., entities taxed at the bank level) and S corporations that are not taxed at the bank level. Without opining as to which measurement is better for the purposes of bank analysis, I will note that using pretax ROA instead of plain ROA for the purposes of my analysis would not have yielded much different relative comparisons among the different UBPR Peer Groups except for the fact that the largest peer group (banks larger than $3 billion) compared much more favorably, presumably because almost no S Corporations existed within that group. Notwithstanding their improved relative performance, Banks in that peer group still finished behind banks in peer group 7 (i.e., $100 million to $300 million, 2 or fewer branches, and with a non-metropolitan main office) with regards to average pretax ROA over the last 5 years (i.e., 0.97% vs. 0.87%).
[3] Paul Kupiec and Stefen Jacewitz, “Community Bank Efficiency and Economies of Scale,” FDIC, December 2012.