Measuring Asset Quality
Safety and Soundness Update - September 2011
Published September 1, 2011 | September 2011 issue
Analysts review many accounting-based metrics to ascertain the condition of a bank. They pay particular attention to measures gauging the repayment on loans and securities (i.e., measures of asset quality); loans and securities make up the vast majority of bank assets, and a drop in timely and full repayment naturally weakens a bank. The so-called Texas ratio—which got its name from its use to analyze weak banks in Texas during the 1980s—is one measure of asset quality that has received attention during the recent downturn in banking conditions.
In lay terms, the ratio compares loans in default plus repossessed property from defaulted borrowers to the financial resources the bank has to absorb losses. The ratio has a straightforward intuition. The higher the ratio, the more loss-producing assets the bank has relative to the money the bank has to cushion those losses. More precisely, the ratio is nonperforming loans plus real estate owned divided by tangible common equity capital plus loan loss reserves. A ratio over 100 percent means that the loss producing assets exceed the cushion. (The developer of the ratio, Gerald Cassidy, viewed 100 percent as a key threshold that only very weak banks crossed.)
What does the Texas ratio tell us about conditions of Ninth District banks (see Graph 1)? First, the vast majority of banks in the district had very low ratios during the 2000s. Second, the share of banks with higher ratios started to decline in 2010 and continued to decline into 2011, suggesting that conditions have begun to stabilize. Finally, the ratio worsened at the same time that banking conditions worsened; the ratio did not seem to predict the crisis.
We explore this final point by dividing all Ninth District banks into satisfactory and less than satisfactory conditions. We make this division based on the safety and soundness ratings of banks made by bank supervisors. Banks receive a 1 through 5 rating under this system, with 1 reserved for the banks in the best condition and 5 for banks in the worst. Supervisors consider a bank receiving a 1 or 2 rating to be in satisfactory condition, while banks with a 3, 4 or 5 are in less than satisfactory condition. We make that division based on ratings as of Aug. 12, 2011. (By definition, this excludes banks that previously failed or merged, but we think this will not materially bias our results, as only a small number of banks fall into this group.) We then review these two groups’ Texas ratios over time. As Graph 2 indicates, the Texas ratios of the banks that ultimately became weak were always higher than the banks that remained in good condition. But the difference in the ratios seems slight prior to the crisis that started June 2007.