Jeff Judy

Jeff's Thoughts - May 12 , 2010

The End of the Story

Every business, whether it's one of your customers or your own bank, has a story to tell. You write a script (called a "business plan") that lays out everything that is going to happen, the sequence of events, and the outcomes that will be achieved by the characters in the story. Your story probably anticipates some twists and turns in the plot, and you write chapters that show how the heroes deal with adversity (their "strategies") and get to those desired outcomes for a happy ending to the tale.

Of course, real life doesn't follow scripts as faithfully as we would like. That's why prudent businesses set aside some reserves for a "rainy day," to give them the means to respond to those additional plot twists the real world is so fond of inserting into our nicely crafted stories.

The banking business stands out from others in our marketplaces because it is in fact required to have that rainy day fund, and I have written about ALLL in recent editions (see "ALLL: Are You Post- or Pre-Cyclical?" and "Is ALLL Well with You?" ). The challenge for us, then, is to set those reserves at just the right level that balances protection from future ugly surprises against availability of funds for current operations.

In other words, if we set our rainy day fund too low, as many banks did before the recession, we may not be able to cope with the way real life rewrites our stories. Set it too high, and we are overpaying for protection, if you will, stashing away resources that we could use to grow our banking business.

At many banks, for many years, setting reserves was an art more than anything else. (Frankly, that is still the case way too often.) But now we have tools for more accurately predicting how our credit stories, at least, will turn out, allowing us to set aside reserves with much more precision.

These tools are based on the understanding that our portfolio story is made up of many individual sub-plots, or "credit transactions," and that every credit transaction, every loan, has its own ending: repayment in full, partial repayment, charge-off. All loans start out healthy (we hope), and then some of them become safer or riskier over time. Within the bank, we see them migrate into better or worse risk ratings, and that has lead to the term "migration analysis," an attempt to anticipate the paths of various segments of the portfolio so we can predict how the portfolio story ends, and prepare accordingly.

I and my associates spend a good deal of time, and more than a few frequent flyer miles, helping institutions develop their own systems for performing this kind of analysis, especially because every bank's story is unique. We want to use your bank's prior loan behaviors, the ratings shifts and repayment patterns you have seen, to predict future loan outcomes. By looking at prior time slices of behavior we can generate an exhaustive view of future outcomes that will have a major impact on your own business success.

But we have found that the term "migration" does not always seem to help our clients embrace this analysis. Migration is a concept that is easier to understand once you are already in the process. In other words, it is only after you have begun the work of tracking your portfolio and your risk categories that you can see how things change.

So we have changed our story, and have begun talking about "terminal risk analysis." Terminal risk is simply how a credit story ends -- the termination of every loan relationship is either repayment (not always full repayment) or a charge-off. Banks that learn to predict how many of their individual credit stories will have happy endings, and how many are likely to have sad endings, will be able to set their reserves right where they have to be to achieve that balance I mentioned earlier.

Every portfolio includes stories that are going to have sad endings for some of your customers. The purpose of terminal risk analysis is to make sure those particular stories don't combine to produce a sad ending for your bank's story as well.