In a recent blog (HERE), we reviewed and analyzed commercial mortgage loans originated in 2017, and we identified some strategies that community banks could deploy immediately in 2018 to increase their return on assets (ROA). We concluded that given the average cost for banks to originate and maintain commercial loans, community banks could increase ROA by approximately 39bps through increasing loan size from the current community bank average commercial loan size to somewhere between $1-2mm in loan size. While some community banks may not be able to increase their average loan size by such a large amount, nonetheless, ample opportunity exists to target larger loans through differential credit spreads and lender compensation. In this blog, we further explore the dynamics of loan size, loans per full-time equivalent employees (FTE), and average industry salary per employee. Again, we aim to identify strategies that are simple to implement, available for most banks, require little to no new investment and can be immediately effective in increasing ROA in 2018.
The Data and Observation
We reviewed the latest available FDIC data for all banks and analyzed loan size, loans to FTE and salary per employee to determine how these variables affect banks’ ROA. Through our loan pricing model and loan trading desk we track individual loan structure, pricing, and ROA to determine where community banks can enhance loan portfolio returns.
Below is a table showing performance for various bank groups divided into asset size. Within each bank group, we measured the correlation between ROA and Loan-per-FTE; and, ROA and Salary-per-FTE. Some of the observations are expected while others are surprising.
There are a number of key takeaways from the table and our summary of the results are as follows:
Loan-per-FTE: As we indicated in our previous blog on the subject, while larger banks originate larger loans, the size difference is not proportional to the size of the balance sheet. The average non-residential commercial loan for all banks in the country in 2017 was $1.40mm, and for national banks that average is $4.1mm, it is only approximately $300k for community banks.
ROA and Loan-per-FTE: In every bank group, there is a positive correlation between ROA and loans per FTE. In all bank groups, except one, that correlation is statistically meaningful. Banks with higher loans-per-FTE demonstrate higher ROA.
ROA and Salary-per-FTE: In every bank group, there is a positive correlation between ROA and salary per FTE. That relationship is surprising since higher salaries per employee lead to higher ROA and are contrary to the paradigm that lowering costs leads to higher performance.
The data we observe can be explained by analyzing the economics of commercial loans. Commercial loans have a few common characteristics, as follows:
- High origination costs,
- Relatively high maintenance costs,
- Low yield,
- Highly scalable revenue, and,
- The high cost of labor as a percentage of all input costs.
Let’s consider a typical commercial loan, investor CRE, 20-year amortization, five year fixed rate at 4.00%, risk rated 3 (approximating 1.8% annual probability of default). Changing only the loan size creates a change in the risk-adjusted ROA/ROE. We ran this loan through a loan pricing model and changed only the loan size. The output from our loan pricing model for four sample scenarios appears below.
The above loan at $2mm translates to a 17.2% ROE (1.7% ROA) while the same loan at a community bank’s average loan size of $300k translates to a 9.9% ROE (1.0% ROA). The difference is mostly driven by the cost of acquisition of the loan versus revenue of the loan. For the $2mm loan, the cost of acquisition is approximately $14,750, the lifetime cost of maintenance is $18k, and the lifetime revenue of the loan is $332k. For the $300k loan, the revenue is proportional at $50k but the cost of acquisition is disproportionately higher at $6,200 and the lifetime maintenance cost of $7k is also disproportionately higher.
The issue is that many smaller banks are focused on net interest margin (NIM) but do not measure ROE/ROA on a loan or relationship level. The additional problem is that the average lender can only handle 15 to 35 relationships (depending on the complexity of the products and customers) and we will write more about this in another blog. Some bankers will attempt to rebut the above argument by stating that they obtain higher yields on lower credits to maintain a profitable loan portfolio. However, our research shows this not to be the case. Banks do not charge enough for smaller credits to make them profitable. We reviewed almost 300 commercial loans originated by community banks across the country in 2017 and considered the relationship between credit spreads (loan pricing) and loan size. The data appears below – many community banks are not pricing smaller loans at higher credit spreads.
Finally, we are stumped on why higher salary-per-FTE leads to higher ROA. We assume that it could be one of two factors: 1) banks that operate in higher salary geographies, and thus must compensate employees to market, also exhibit higher ROA because of those geographies; and/or, this is more likely, 2) better employees are better compensated, and better employees can earn higher ROAs.
There is a strong correlation between loans-per-FTE and performance. While community banks need not increase their average loan size to match larger competitors, increasing average loan size does drive ROA. As the data above indicates, loan size to optimize ROA for community banks is the $1-2mm range. Higher compensation also correlates to bank performance. Obviously, banks should not be motivated to increase salaries for the sake of increasing ROA – rather, better employees do lead to better performance.
Finally, our pricing data shows that the market for small commercial loans is inefficient and many banks often underprice credit given the economics of loan origination, maintenance, and pricing. If your bank faces this type of competitor, then your bank needs to invest in bringing your cost of origination low enough so you can book the loan at the below market price and still achieve your cost of capital. Alternatively, you can plan on booking that loan as a loss leader, but you need to make sure your bank has the training and the execution to decipher which loans lead to profitable relationships in the long run – a feat that is difficult to pull off.
Submitted by Chris Nichols on January 16, 2018