Weak credit conditions in the banking industry have many looking over their shoulder, turning over rocks and digging through the credit portfolio to uncover unforeseen risk. We have recently written in this column about the risk of increased loan defaults and market volatility, so today we focus on the interest rate side of the fence to discuss problems lurking in the shadows.
Many community bankers originate 10Y maturity loans by breaking the cashflows into two 5Y pieces and pricing the reset in 5Ys off the "then" 5Y Treasury plus a locked-in spread (or they use Prime). Most lenders are leery of originating a 10Y loan, so by breaking it into two 5Y pieces, a psychological comfort zone is maintained, customers remain relatively happy and many feel the risk of the transaction has been reduced.
Let's examine this concept further to see what risk might actually be lurking in the loan portfolio. To begin, we assume a bank originated such a loan exactly 5Ys ago and that the customer did not prepay it. The loan was originally priced at a spread of 5Y Treasuries plus 250bp. Back then, the 5Y Treasury yield was a low 2.75%, so by adding 250bp, the bank earned an overall coupon of 5.25%. Assuming a loan size of $5mm and no credit losses on this loan, the first 5Y period resulted in interest income to the bank of $1.3mm.
Let's say our example loan reset yesterday for the upcoming 5Y period and the customer locked in a rate of 5.88% all-in. While it remains to be seen whether or not that will be a good situation for the bank, consider some of the risk components in this transaction.
By choosing Treasuries as an index banks are inherently selecting a risk free rate as a starting point. Because of this, the loan does not directly take into account systemic problems in the credit market. Put another way, while Treasuries can be whipped around wildly as a result of credit issues in the marketplace, it is risk-free by definition, so spread is not embedded in the rate. While banks capture the credit risk of their borrower at a point in time (i.e. 250bp over current Treasuries), they do not capture the risk of the general marketplace. Normally this is not an issue, however in times of extreme volatility it can become material.
Since the beginning of the year, 5Y Treasury yields compared to swap spreads have fallen roughly 130bp. That means banks with loans that reset against Treasuries as an index are under pricing their risk by as much as 30%.
While there is no perfect solution to this dilemma, one thing community bankers should consider is to build in covenants in loan documents to include a market credit risk component. Another way to handle this is to base the reset rate off the swap curve to incorporate a credit component.
Regardless of which method is selected, banks should try to get a better handle on the interest rate risk exposures in the loan portfolio. We suggest forward testing to "pretend" all loans up for reset next year actually do so at their current spreads. Depending on how much is rolling over, the impact on NIM will probably be around a negative 100bp. By running that through the ALM model, bankers can get a better handle on how much work will need to be done on the deposit side to counterbalance these effects.
We urge banks to take the time to calculate how much "basis risk" exposure they have in the portfolio. As a quick reminder, basis risk refers to the relative movement between one index and another. For example, community banks that originate loans based on Treasury rates (or FHLB), but fund themselves with deposits based on Libor (or FHLB), are exposed to the risk of unexpected changes in these spreads.
Current market volatility provides a real-life example of why community banks should enhance they way they measure, monitor and manage these risks to help protect the banking franchise over the long-term. We urge community banks to get a jump on this right away in an effort to move the risk out from the shadows and recognizing the exposure for what it is.