« Margin-Squeeze Predictions Starting to Bear Out - American Banker Article | Main | Stress Testing Best Practice: For Good Results, Keep It Simple »

Comments

Gary Brinley

Brad,
We have seen some commentary from our examiners related to Net Earnings at Risk. We have been outside our risk limits for a number of years, with it worse more recently due to higher provision expenses (similar to the bank you talk about in your post). We are talking about changing our Net Earnings at Risk limits. What policy limits to you see from banks in your universe?

Brad Olson

It used to be that the rule-of-thumb Earnings at Risk (EAR) limit was -15.0%. If you look back in history at the average level of EAR you see it hovered between -14.0% and -20.0%.

http://www.olsonresearch.com/IRData/2004Q4/interestraterisk/interestraterisk_c/default.htm
http://www.olsonresearch.com/IRData/2007Q4/interestraterisk/interestraterisk_c/default.htm

It wasn’t uncommon to see banks (especially smaller ones) with ALCO EAR limits of -30.0%.

But that was back before the current credit crisis.

Now, with much higher levels of provision & overhead expense, average EAR has increased substantially reaching nearly -40.0% by 4th quarter 2010.

http://www.olsonresearch.com/IRData/2010Q4/interestraterisk/interestraterisk_c/default.htm

A -40.0% ALCO limit for EAR is not that unreasonable. Especially if as an industry we’re going to continue with a high level of provision expense. If your EAR measurements are falling outside -40.0% I’m guessing it’s due to abnormal levels of provision and other non-interest expense. If so, document the anomaly along with what you expect a more “normal” level of expense to be.

If the elevated overhead levels are not an anomaly I think there may be more than ALCO policy to worry about. Such levels of overhead are probably not sustainable.

The comments to this entry are closed.