BID® Daily Newsletter
Mar 7, 2007

BID® Daily Newsletter

Mar 7, 2007

LEMMING TESTING


It is interesting that many studies show while 67% of small businesses will survive past the 2Y mark, only 44% will make it as long as 4Y. Businesses aren't lemmings, however and grouping them all into such broad buckets can be quite dangerous. This is important not only because Independent bankers make lots and lots of loans to small businesses, but also as we look at the behavior of such businesses when interest rates begin to shift around. Given the flat nature of the current yield curve, getting a better handle on the some of the quirks of our customers' behavior is a worthy endeavor. This brings us to an old saying in bank asset-liability management ("ALM") that says it is always the next 100bp that hurts the most. You see, human behavior is a bit quirky and it is certainly not symmetrical. As one ponders that, consider that businesses do not act the same way when rates go up 100bp as they do when rates drop 100bp. In addition, businesses do not act the same way the first time rates rise or fall 100bp, as they do when rates rise or fall the next 100bp. In fact, on the way toward an up 300bp scenario, the first 100bp accounts for 53% of the duration shift for bank fixed rate assets, while the remaining 100bp shifts make up the rest. Consider that the lower the rate on the asset, the more interest rate sensitivity it exhibits. Put another way, a loan originated at 6% will be less likely to prepay when rates rise to 7%, than a loan originated at 7%. Another point to keep in mind is that with each passing day, the percentage of a bank's assets originated at "current" rate levels increases (particularly the longer the curve remains flat). Looking back over the history of banking, there have been few times when the dispersion of individual interest rates has been so closely grouped together. Just as diversification helps with credit, so it does with rates. Closely grouped asset and liability rates increase interest rate sensitivity. One quick way to check this on an ALM model is to see what the duration difference is between up 100bp and down 100bp rate shocks. If these are similar, either your bank has originated and purchased assets without optionality (highly unlikely), or the model is not calibrated properly. Second, compare the duration difference in the up 100bp and up 200bp scenarios. Subtract the duration from each other. If these numbers are the same, find out why. Very few assets will react in parallel fashion in the real world. Remember, that when measuring bank performance, extension risk has historically presented a bigger problem for independent banks than contraction risk. Just because we have a flat rate curve now does not mean it will stay that way forever. Having a poorly calibrated ALM model gives the bank a false sense of security, while leaving the door open to a future of potentially higher levels of underperforming assets. Small business customers are not lemming and they certainly do not behave the same way all the time in all rate environments. Understanding this is the first step to enhancing the robustness of the ALM modeling process.
Subscribe to the BID Daily Newsletter to have it delivered by email daily.