Brunini Law

In April, 2008, I personally needed a change of scenery. I was extremely lucky to be part of arguably the largest banking law practice in the state at that time, and I was equally fortunate to have arguably the state’s best known banking attorney as my mentor. However, the pressures and stresses of balancing the practice of law, billable hours, and my family, which was four months into our second child, was beginning to wear on me. As a result, I made a decision, maybe somewhat hastily, to leave the practice of law and join the bank back home that my family had been involved in for some time. At that point in my life, “Bankers’ Hours” was a dream that was hard to forget and impossible to pass up.

Shortly after I left the firm and joined the bank, an older banker whom I had worked with and have a tremendous amount of respect for called me to chastise me for not telling him that I was leaving the firm. This banker had been around the business for a long time and had seen a lot. He had the charisma of a Marlon Brando with the down home common sense of an Andy Griffith, even though it is safe to say he did not have the Hollywood looks of either. After properly reprimanding me, he turned the conversation to wishing me luck and giving me whatever nuggets of wisdom I could use in my newfound career. He ended the conversation with this statement that has stuck with me since that day: “This business ain’t hard, son, you just can’t make bad loans.” Ironically, five months later, all hell broke loose in banking and within our nation’s economy, and his words remained in the background of my psyche like a continuing subtitle that I used to interpret the chaos that was ensuing all around the banking industry.

In December of 2012, the Federal Deposit Insurance Corporation (“FDIC”) conducted its “Community Banking Study,” which was “a data-driven effort to identify and explore issues and questions about community banks.” It analyzed banking data over the period of time between 1984 and 2011 and tried to analyze trends and circumstances that were altering the banking landscape for community banks and which impacted the performance of community banks as they entered the “Great Recession.” Chapter 5 of that study was a comparative performance of community bank lending specialty groups, and it focused mostly on how community banks had shifted their lending strategies prior to the recession and what affect that strategic change had on their performance during the recession. I found the following points to be particularly interesting:

• In 1984, retail loans represented over 61 percent of all loans at “community banks[1]”, compared with 35 percent of all loans at “noncommunity banks.” By the end of 2011, these ratios had basically flipped so that retail loans made up 36 percent of community bank loans and 54 percent of noncommunity bank loans;
• Between 1984 and 2011, residential real estate loans fell from 47 percent of community bank total loans to 32 percent, while commercial real estate loans rose from 21 percent of loans to 42 percent;
• Commercial Real Estate (“CRE”) specialists[2] experienced the most volatile earnings performance during the examined period, and their pretax ROA trailed the community bank average by more than one-third;
• CRE specialists had a high frequency of failure during the period examined, failing at a rate that was more than two time as frequent as similarly situated community banks;
• Between 1991 and 2007, the number of CRE specialists increased fivefold, going from less than 4 percent of all community banks in 1991 to almost 30 percent of community banks at their peak in 2007;
• For the entire study period, community banks with Commercial and Development (“C&D”) loans, a subset of CRE loans, greater than 10 percent of their assets were 2.8 times more likely to fail than the average community bank, while those with C&D loans below 10 percent were less likely to fail than the average community bank;
• CRE specialists were the worst performers over the entire study period, performing slightly better than the average for all community banks in good economic times, but performing significantly worse during periods of banking crises; and
• CRE specialists were primarily headquartered in metro counties (80 percent), and 74 percent of community banks with at least 10 percent of assets in C&D loans were primarily located in metro areas.

When I was examining the UBPR data for the period of time that mostly followed the FDIC’s study period, my first hypothesis for explaining the earnings discrepancy between metropolitan and non-metropolitan banks was that it had to be related to these “bad” CRE and C&D loans cited in this study that the wise banker warned me about. It seemed quite logical to me that metropolitan banks would have much lower earnings, even during this “recovery” period, because they were still fighting loan provisions and non-accrual loans from CRE and C&D loans that were just more prevalent in their markets.

After testing this assumption, it did appear to hold up, at least partially. From the beginning of 2010 through the end of 2014, banks under $300 million in assets and in a metro area averaged loan provisions that were 0.28% of total assets compared to loan provisions that averaged 0.17% of total assets for similarly sized banks headquartered in non-metro areas. Likewise, the same metro banks had average non-accruals equal to 1.72% of total loans during that period, while their rural counterparts had a non-accrual percentage of 1.04%. There is no doubt that higher balances of “bad” loans are still holding metropolitan banks back. By the way, for what it is worth, banks larger than $300 million averaged provisions and nonaccruals that were higher than each of these averages, even though their average earnings were on par with the smaller, rural banks and much higher than the smaller, metropolitan banks.

However, when I considered that rural banks under $300 million in assets had earnings that basically doubled those of similarly sized metro banks from 2010 through 2014, this answer alone did not seem to be sufficient to answer the question entirely, and it wasn’t. The rest of the story became evident when I dug deeper in the bottom half of the income statement, and that will be the topic of my next article.

[1] The study 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.”

[2] In order to be considered a “specialists” for the purposes of this study, a bank had to hold loans of that type greater than 33 percent of total assets.

“Wait a minute,” you say, “this can’t be right. Everything I have heard in banking for the past 20 years has told me my bank must get larger. We must abandon these old, rural markets for the greener pastures of America’s cities and suburbs, where loans and deposits flow like milk and honey.” Well, sit down, because the revelations discussed in my last post become even more dramatic. As mentioned in my earlier post, the UBPR divides the 12 Peer Groups covering banks under $300 million not only based upon size, but also based upon their “metropolitan” or “non-metropolitan” locations and their number of branches. By examining these 12 peer groups, I discovered that ”metropolitan bank” peer groups had average ROAs over the last 5 years (i.e., 0.44%) that collectively averaged approximately half those of “rural” peer groups (i.e., 0.87%), while having collective average ROEs (i.e., 3.50%) that were less than half of those recorded by their more rural cousins (i.e., 8.02%). Not only are rural banks more profitable than their metropolitan competitors, they have been running circles around them for the last five years. How do those same rural peer group averages compare to the averages for the “larger” peer groups 1, 2, and 3? Very favorably, thank you. The collective average ROE (i.e., 8.02%) for these rural peer groups (i.e., 5, 7, 9, 11, 13, and 15) was higher than the average ROE for each of the three larger peer groups (7.79%, 7.60%, and 7.59%, respectively), and the collective average ROA (i.e., 0.87%) for these rural peer groups was higher than all but one of the average ROAs of the three larger peer groups (0.88%, 0.79% and 0.80%, respectively). Therefore, maybe it’s not size that is the ultimate differentiator, but instead location. Contrary to everything I have believed and heard to this point, maybe it is a bank’s rural location that is its ace in the hole!

Chart 3: Average “Metropolitan” ROA vs. “Non-Metropolitan” ROA, 2010 – 2014

Chart 4: Average “Metropolitan” ROE vs. “Non-Metropolitan” ROE, 2010 – 2014

What is the one rumor that appears to be true? Well, that may be that brick and mortar are the albatross that everyone believes them now to be. Banks in the last 12 peer groups that have “more” branches (i.e., 4, 5, 8, 9, 12, and 13) had a collectively lower average ROA (i.e., 0.56% to 0.75%) than the peer groups with “fewer” branches, and their collective average ROE was lower as well (i.e., 5.20% compared to 6.33% for peer groups with ”fewer” branches).

Chart 5: Average “More Branches” ROA vs. “Fewer Branches” ROA, 2010 – 2014

Chart 6: Average “More Branches” ROE vs. “Fewer Branches” ROE, 2010 – 2014

Maybe I’m the only one surprised by these revelations. After all, the “User’s Guide for the Uniform Bank Performance Report – Technical Information” notices the following:

Consistent differences in peer group performance are apparent over time. For example, the average non-branch bank in a non-metropolitan area tends to have lower overhead, lower noninterest income, higher profitability and higher capital ratios than similar sized branch banks located in metropolitan areas.

Something tells me that not many people pick this up for bed time reading, though, so its truths may not be widely known. “Still,” you say, “how could it be that smaller, more rural banks turn a higher profit than larger, more metropolitan ones? That can’t be true since my bank is dedicating most of its resources to the nearby metropolitan area where all of its loan growth has occurred over the last five years.” I will elaborate further on what the UBPR data reveals as reasons for this irony in my next post, but to give you a hint, it is clearly related to the first metric listed in the “User’s Guide” quote above ( i.e., lower overhead). You may be generating a lot more loans and deposits from that metropolitan area and growing quickly as a result, but it also costs you a lot more money to do so. Maybe your father knew what he was telling you when he preached that a penny saved is always a penny earned.

We all know the story and ultimate conclusion, right? Commercial banking is becoming more complex, regulations are multiplying exponentially each passing day, commercial activity and related loans are migrating more and more to metropolitan areas, so therefore small town, community banks are quickly becoming a thing of the past. They simply can’t keep up. Not only are they losing loans to banks in faster growing, more populated areas, but they also are struggling to hire sufficient staff to comply with regulation, much less chase loans. They just don’t have the economies of scale or markets to support the inevitable overhead explosion and attract the top talent to their slow growing, rural economies. Sadly, their days are numbered. Within a decade or two, rural community banks with assets less than $1 billion will go the way of the dinosaur, victims of that terrible meteor named Dodd-Frank.

This must be the truth, right? Everything we learned in macroeconomics demands it. Bigger banks have more economies of scale to support the rising fixed costs associated with commercial banking, and smaller banks will eventually tap out. Metropolitan banks have more access to capital, loans, and growing deposit bases and therefore have an inherent advantage over their rural cousin the country bank. As a former CFO and COO of a rural, community bank, I know I bought it hook, line, and sinker. Looking at growing overhead, shrinking loans, and aging management, I was convinced this storyline was the only possible one. Heck, it’s a big reason why I decided a future in community banking was not for me, opting instead to return to the much more stable world of practicing law (note the presence of my tongue which is firmly implanted in my cheek).

The problem is, five years after the end of the great recession, the bank statistics published in the Uniform Bank Performance Reports (“UBPR”) by the Federal Financial Institutions Examination Council (“FFIEC”) simply don’t support this idea that small town community banks are dead, or that they are even dying. As a matter of fact, there is some argument to the contrary, at least in certain contexts.

In reaching this conclusion, I examined the UPBR average peer group data from the last five years for each of the 15 major peer groups for insured commercial banks. I excluded the peer groups related to De Novo banks created in the last five years which had the potential to skew the analysis due to the unique challenges faced by De Novo institutions. These 15 different peer groups, which included 5,619 banks, are delineated as follows:

As of December 31, 2014, the average number of banks per peer group was 374.6 banks. The largest peer group by far was Peer Group 3 (i.e., banks between $300 million and $1 billion), which included 1,254 banks. The smallest peer group was Peer Group 12 (i.e., banks less than $50 million located in a metropolitan area and having 2 or more full branches), which contained 63 banks. The median peer group was Peer Group 2 (i.e., banks between $1 billion and $3 billion) with 321 banks.

This UBPR data separates banks into peer groups that distinguish them not only on the basis of size, but also based upon the number of full service branches operated by a bank as well as whether the bank is located in a metropolitan or non-metropolitan area. For the purposes of clarification, it is important to note that a bank may be classified as a non-metropolitan bank and still have full service branches in a metropolitan area, and vice versa. For example, a $150 million commercial bank whose main office is in a non-metropolitan area but who also operates another full service branch in a metropolitan area is part of Peer Group 7, which includes banks between $100 million and $300 million of assets that have 2 or fewer branches and are not located in a metropolitan area. Therefore, it is the main office location of the bank that controls and not the location of its branches. For the purposes of the UBPR, a metropolitan area is one classified as a Metropolitan Statistical Area by the Office of Management and Budget.

I first stumbled upon the truths presented by the UBPR while I was analyzing the performance data of a client. With the same prejudices in mind that I stated in the opening two paragraphs, I decided to compare that bank’s data to statistics for banks in “larger” peer groups. What I discovered astonished me and interested me to the point that I decided to dig deeper. Not only did the bank’s statistics compare more favorable to the data of “larger” peer groups than it did to statistics of its own peer group, but the average numbers for the bank’s peer group 7, which is assigned to relatively smaller, rural banks, seemed to soar well above some of its larger, more metropolitan cousins.

Peer group 7, which, as mentioned above, is reserved for banks between $100 million and $300 million in assets that are located in a non-metropolitan area and have 2 or fewer branches, averaged the highest Return on Equity (“ROE”) (i.e., 10.47%) and Return on Assets (“ROA”) (i.e., 1.16%) of any peer group over the last five years. What about the next highest ROE and ROA? Well those belonged to “rural, community” banks as well. Peer Group 11 (i.e., $50 million to $100 million, 2 or fewer branches, and non-metropolitan area) boasted the next highest ROA of 0.99%, and Peer Group 5 (i.e., $100 million to $300 million, 3 or more branches, and non-metropolitan area) claimed second place in ROE with 9.12%. Third place in each category also belonged to Peer Groups 11 and 5, just in reverse with respect to the category. Peer Group 1, the peer group for the nation’s largest banks (i.e., more than $3 billion in assets), doesn’t show up on either list until you look down to fourth place, where it finished with an average ROE of 7.79% and an average ROA of 0.88% over the last five years. In my next post, we will start to examine what could be the explanation of this and what secrets it could reveal to you regarding operating a community bank in this challenging environment.