The cliché of "What keeps you up at night?" can be answered in several ways. Since sleep is so valuable, it would be nice to know if there is any statistical support for your tossing and turning. To find out, we did some analysis tht seems to indicate we have more banks on the verge of failure than at any time since 2002. The practical question is, if your bank is going to fail, what are some early warning signs? To answer the question, we looked at all bank failures since 1989 and excluded those that failed due to internal fraud.
The first thing that jumped out is the fact that thrifts made up almost half the failures over that time. To derive more actionable results, we excluded those institutions and ran a multiple regression analysis looking at a variety of variables. A couple of interesting points emerged.
For starters, if you are located in TX, LA, OK, FL or CA; your probability of default is about 5% higher than the rest of the country (a 6bp higher probability of default). This is due to a variety of reasons, but competition and a high correlation to real estate/energy industries likely explain most of this. Other states exhibit fairly random defaults indicating there is little correlation outside the specific states mentioned.
Getting even more granular, concentrations in CRE (those above 400% of Tier-1 capital or more than 40% of total assets) are the single largest reason banks fail by almost a factor of 2. Concentrations in other sectors that also present a quantifiable element of risk, in order, include C&I, consumer, and agriculture.
In fact, concentrations explain 38% of bank failures. The single best thing banks can do to ensure survival is to strive for diversification within the balance sheet - even if it means sacrificing profitability. This is one reason why bankers are reexamining how important a liquid investment portfolio is. A securities portfolio that composes 20% to 30% of assets adds material diversification. The more diversified your loans are, the smaller the investment portfolio required.
Next to concentrations, profitability explains 25% of bank failures. No surprise, net income is the largest single driver, followed by net interest margin.
Explaining 13% of bank failures, capital levels have the third largest influence. This is a bit of a surprise, as prior to the late 1990's, capital was thought to be most accurate predictor of failure. Here, equity as a percent of total assets is the number one ratio that matters, followed closely by allowance for loan and lease losses as a percent of total assets. If you lack diversification and profitability, shrinking the bank is the best tactic to keep from a catastrophic default.
Interestingly, credit quality, as judged by the percentage of non-performing loans to assets, ranks 4th and accounts for 12% of bank failures. Note that similar to investment management, it matters less how you underwrite loans than what you underwrite. Over time, asset allocation is almost 3x more important than the quality of a bank's underwriting standards. Think of the amount of resources banks allocate to credit committee and compare that to how much time is spent setting allocation targets at the strategic level. Most banks spend less than 8 total hours per year setting asset mix, yet this is the largest determinant of credit quality.
Finally, the amount of liquidity on the balance sheet has a 9% correlation, while asset growth has less than a 2% effect.
The takeaway here is that while bank management is charged with profitability and preservation of capital, they are not mutually exclusive goals. As history has judged, bankers that have a well-constructed balance sheet and profitable operations are those that tend to sleep through the night.