BID® Daily Newsletter
Apr 27, 2009

BID® Daily Newsletter

Apr 27, 2009

BRINGING BANK FAILURES INTO PICTURE


Some things we take as given. For example, most of us require "picture-in-picture" when we purchase a television set. Seriously - have you ever seen any one use this feature? While it might be a neat feature in concept, the reality is that most of us can't remember how to use it, can't see the small second picture in the corner, don't have the concentration level to watch 2 shows/games at once and can't find 2 things we want to watch anyway.
The assumption that you need to spend the extra $100 on this technology is similar to the assumption of what causes bank failures. While we know that bank failure is a combination of deteriorating credit quality, capital and liquidity, what causes these problems? In other words, is there a way to predict failure?
We are sure that there is an answer to the question, but after looking at bank statistical data for the years 2004 thru 2008, we can't find anything of statistical significance. Now, we bring this picture into focus, because if you ask some examiners, they will imply that bank failure has to do with concentrations, high cost of funds, brokered deposits, commercial real estate or high overhead. While all this is true, this is like saying that television causes bank failures. While American Idol certainly isn't helping productivity, we doubt the FDIC is looking into it.
That last analogy may not be entirely fair, but it's close. If you look at historical bank performance, take each independent variable and look at its relationship to performance, you end up with a very grainy picture. For example, one of the best predictors of failure is high overhead costs. However, this can explain only 10% of performance. Looked at another way, 9 of 10 failing banks didn't have an issue with overhead costs, so what good is that statistic?
After that, concentrations in CRE explain 6% of the causation for all failures, as does leverage and the amount of capital a bank has. If you lump cost of funds, FHLB advances and brokered funds together, this helps explain only 4% of the causation for failures. Concentration in construction explains 3% of the outcome, while the level of ALLL explains less than 2%.
The statistical implications of this research put a basic tenant of bank performance in high definition - there are many different ways to succeed or fail in banking. Most likely success or failure is a combination of many different factors. To penalize banks for any particular characteristic usually doesn't make sense. Further, in some cases, the regulatory imposed solution ironically, is a greater predictor of failure. While one can argue that some of these banks would have failed anyway, in some cases mandated
asset sales, forced capital raising or restricting liquidity produced a greater indicator than the problem these solutions tried to fix.
To be clear, our point isn't that CRE, cost of funds and the amount of ALLL don't matter, because they do in totality. However, our point is that we need to keep performance in perspective. To prove this point in closing, let us open up one last picture. The largest single factor that we have found so far the projects failure is - what state you are from. Where you are domiciled can account for 14% of failed performance. AZ, FL and CA have a statistically high failure or potential failure rate, much more than any single characteristic that we have looked at so far. Picture the implications of that.
Subscribe to the BID Daily Newsletter to have it delivered by email daily.