Jeff Judy

Jeff's Thoughts - September 11, 2013

FASB Standards Update: The "When" of "Expected Losses"

Preparing for possible credit losses has traditionally been a backward-looking affair. Often the biggest driver of the educated guesses that translate into ALLL and Capital is the state of the credit portfolio over the last couple of quarters. In other words, we look at the losses incurred, and also at the obviously weak loans that seem to be going downhill, and we set aside resources to handle those losses if they all come to pass.

As I have mentioned in a couple of recent articles, we are now facing an FASB Accounting Standards Update that explicitly changes the required language from "incurred loss" to "expected loss." (See "FASB Standards Update: The Meaning of 'Expected'".) As regulatory bodies will require this updated standard at all banks, regardless of size, that simple, single word change has dramatically flipped reserving practice. Talking about what happened recently will be of little avail, in discussions with regulators, as you manage ALLL and Capital. If you can't tell a good story about what you expect in the future, and why you expect that particular future, you are probably in for a regulatory bruising.

In essence, this is a very strong nudge to get less reactive, to become more proactive, or "procyclical," as it has been called. But the phrase "expected loss" raises a related question:

Expected when? Is the loss expected next quarter? Next year?

Even banks that take a more forward-looking approach to forecasting portfololio quality often think in relatively short time frames. It is rather like the headlights in your car, and how far ahead you can see that deer crossing the road. Some banks will drive with their headlights on "high", others will drive with their headlights on "low", but most of them will have a set "distance" in mind for forecasting losses.

Reserves are internal "insurance" based on seeing a deer in your headlights, whatever distance that is.

The problem with that approach is lag time. Just as driving in the countryside with your headlights on "dim" may mean you don't have time to stop your car before getting intimately acquainted with that deer, looking a few quarters out works all right for the more predictable credits, but causes problems for the ones that deteriorate quickly. If you have a credit that has been rated quite good for years, but that suddenly goes downhill -- and we all have those in our portfolios -- you don't have time to build up reserves and protect capital when the borrower takes a dive.

Now the regulators will demand a "life of portfolio" perspective. The credit portfolio assessment model I prefer, Terminal Risk Analysis, looks at pools of dollars in the portfolio and predicts their status when the credit terminates. And for any given dollar in the portfolio, the only terminal status entries we will ever make are "repaid" or "charged off."

In other words, imagine you had a device in your car that could tell you if a deer will step out in front of you any time between the start and the end of your trip -- well beyond the distance illuminated by your headlights.

When we manage reserves over the life of each credit, we smooth out the swings in ALLL and Capital. We don't end up scrambling to build reserves when trouble arrives, and we don't lock away resources in reserves that could be better used elsewhere, just because we recently had trouble.

As a trainer and as a consultant, I've been strongly urging banking staff to start getting their heads around this "expected loss" perspective immediately. As you do that, spend a little time thinking about how the "when" of "expected" plays out in your credit portfolio assessments.