There is a common perception that community banks’ performance improves when short-term rates rise. The analysis of this thinking is especially germane now that Federal Reserve Chair Yellen has clearly signaled that the Fed is ready to make a move by year-end. Since we have debated with so many analysts and bankers over the years, we pulled the community bank data for the last 20 years and present balance sheet composition for the industry. Unfortunately for all of us, there is no empirical evidence to suggest that community banks perform better when rates rise. Historically, interest rate movement had no impact on bank performance as measured by net interest margin (NIM). We conclude that this next interest rate cycle may prove to be painful for many community banks unless bankers take steps to correct duration mismatch.
The graph below plots Fed Funds rates and average NIM for all banks between $100mm and $10B in assets for the last 20 years. The data is clear: there is no relationship between interest rate movement or absolute level of rates and NIM. NIM has been in a slow and steady decline from 4.50% in 1994 to its present level of 3.67%. Rising or falling short-term interest rates have had no effect on NIM.
The traditional thinking has been that as rates rise, bank assets reprice at a faster rate than bank liabilities. This conclusion assumes that banks are asset sensitive. However, this assumption is incorrect because, for this peer group, the correlation between short-term rates and cost of funding is 0.91. Therefore, at least for community banks, liabilities reprice very quickly and this was the subject of a recent previous blog (HERE).
Factors Driving NIM
We believe that average NIM is the most direct and appropriate measure of performance when assessing the impact of rising rates. Unlike return on equity (ROE), return on assets (ROA) or risk-adjusted return on capital (RAROC), NIM is not influenced by credit quality, overhead costs or fee and ancillary business factors. Furthermore, while there are substantial complexities in measuring duration, convexity, DDA betas and correlation coefficients, predicting NIM creates a clear and telling picture of a bank’s earnings ability.
Since we have discussed in a previous blog that community banks’ COF rise in almost in lock step with short-term rates, to predict a bank’s future NIM, we now need to assess the repricing speed of the asset side of the balance sheet. The graph below shows the ratio of long-term assets to total assets. Long-term assets are generally defined as those that reprice in 5-years or more. For our group of banks (4,439 total), for the period from 1994 to 2014, the ratio of long-term assets to total assets has increased from 12.1% to 32.3%. The increase in this ratio is especially rapid from 2006 to 2014, where the ratio almost doubled in 8 years.
This ratio shows that since the start of the last recession, when short-term rates headed to zero, that community banks have dramatically increased the duration of their loan and securities portfolio. We believe that the main reason for this increase is management's desire to preserve NIM by extending asset duration.
While previous cycles have demonstrated that rising or falling interest rates have had little impact on community bank NIM, we believe with the recent extension of asset duration by community banks that rising short-term rates will have a negative impact on NIM for most banks. We further believe that this NIM contraction will be more severe than expected because of the dynamic nature of the loan portfolio. Short-duration loans will not reprice as expected, but will instead refinance into longer-term assets – exerting additional pressure on banks’ balance sheets. Also, floating rate loans with floors will behave like fixed rate loans through the initial rate rises.
We have been running models to predict NIM behavior for banks across the country. Our model uses call report data and, therefore, may not be as robust as what most banks currently use for ALM purposes. However, our model is simple to run, and can accommodate various what-if scenarios such as changes in future expected rates, different deposit betas, loan mix assumptions and hedging strategies. The output looks like the graph below. Please let us know if you would like to see your bank’s NIM forecasted for 3 years and some recommendations on strategies.
Community banks’ cost of funding has historically been very closely tied to short-term interest rates. With short-term rates expected to rise in the near future, we considered banks’ asset duration. Asset duration is at its all-time high with long-term assets to total assets at 32.3%. If rates rise as expected – based on the futures market this is expected to be slow and steady – most community banks will experience NIM contraction and pressure to refinance existing short-term assets into longer terms. However, the additional concern is that if rates rise faster than expected and surge deposits flee, then some banks will be in a dangerous position of having to preserve margin to remain profitable. Banks should take immediate steps to develop a strategy to preserve their NIM as short-term rates rise over the next 12 months.
Submitted by Chris Nichols on July 15, 2015