In Pursuit of Return on Equity

When performing ratio analysis to determine a company’s profitability we should remember that a ratio has at least two data points: a numerator and a denominator. It doesn’t matter if it is banking, retailing, or widget making.
In banking, the standard profitability ratio has long been return on equity (ROE). That is… until the financial crisis of 2007-08. It was in the aftermath that capital, the denominator in the return on equity calculation, resumed its place as king.
Where has the pursuit of ROE led us? Yes, it made us focus on profitability, the numerator in the equation. But it also resulted in us looking at bank equity. The smaller the E, the better the ROE, right?
Bond salesman loved the concept. They encouraged their bank clients to borrow from their respective Federal Home Loan Banks (FHLB) or chase high cost deposits to fund the bonds they sold for minuscule spreads. The logic: blow up the balance sheet, eek out incremental profits, and reduce the E. Genius! And equity analysts, who coincidentally worked for the same firms as the bond salesman, loved it too.
During the period 2002-07, when loan growth outpaced the ability to fund it, FIs took on more FHLB borrowings and high cost deposits. This was a higher spread concept, because loans typically had greater yields than bonds. Loans also typically had greater credit risk. FIs did provision for such losses, but accounting rules and how the SEC enforced them did not allow FIs to “over-reserve”, whatever that means. This was determined by the high profile case the SEC brought against SunTrust in the Fall of 1998 for managing earnings through the loan loss provision. Read a summary of it from the Atlanta Fed here. The Fed research piece on the subject, written in 2000, is almost comical to read given what has happened.
So, in pursuit of ROE, banks leveraged up their balance sheet. They thinned their capital leaving them more susceptible to distress during economic hard times. This distress was clearly evident when I recently performed some “where are they now” research on the highest performing ROE FIs in 2006, the last normal year prior to the crisis (see table).
A note on the data. I searched all publicly traded banks and thrifts that existed in 2006 and sorted them by the highest ROE. But I also ensured that their prior year ROE was similarly high, in order to weed out one-time gainers. In other words, I wanted consistent, high ROE FIs.
The results are telling. Of the top 10 ROE FIs of 2006, only three are currently profitable. Two are under a regulatory agreement. The remaining five failed. Yes, failed.
How does an FI, sitting atop the ROE chain, fall so far so fast?  As mostly always the case, it was bad loans. But many if not most FIs have had loan troubles. What made so many in this group take the perp walk to the FDIC? I already mentioned their inability to put extra away for a rainy day via the loan loss provision. But the pursuit of ROE encouraged levering up the balance sheet to leave little in excess equity to absorb the losses. Inadequate loan loss allowance, relatively low equity. The recipe for disaster in bad times.
I think ROE will re-emerge as ONE important indicator of profitability. But I don’t think we will make the same mistake twice. Having the past three years as history in our loan loss allowance calculations, our regulators and the SEC will probably permit us to be more aggressive in provisioning. Needing the federal government to chip in capital because we did not have enough to withstand the storm is resulting in higher capital requirements, and lower ROE expectations. Analysts take note.
What do you think should be the primary measure of FI profitability?
~ Jeff