Trepp Sees Slight Drop in At-Risk Banks

NEW YORK CITY-Trepp LLC says its updated Trepp Capital Adequacy Stress Test results show a marginal quarter-to-quarter improvement in the percentage of banks that would be at risk of going under amid severely adverse economic conditions. The latest T-CAST analysis shows that based on third-quarter 2012 numbers, 747 banks out of 6,000, or 12%, would fail the stress test without raising additional capital, compared to 12.7% in Q2 of last year.

“While conditions in these states have improved, the high T-CAST failure rates indicate that more progress in capital raising and balance sheet healing is needed before these areas fully emerge from the recent failure cycle,” Matt Anderson, Trepp's managing director, says in a statement.  The improvement can be attributed to a quarterly accrual of retained earnings and the fact that the official inputs that were formally released by the regulators in mid-November were not quite as onerous as Trepp had forecast.

Further, Trepp says that on a state-by-state basis, the increases in fail risk tend to be smaller than improvements in this area. For example, North Dakota showed a 12.79% decrease in its fail risk, while Florida's increase in fail risk was only 0.94%. Nonetheless, the Sunshine State registered the largest failure rate of all at 31.75%, followed by Georgia at 29.19% and South Carolina at 24.56%.

Among large states, those with the highest T-CAST failure rates included Tennessee at 17.9%, Illinois at 15.7% and Minnesota at 15.2%, in addition to Florida and Georgia. “These are also the states that have experienced the highest rate of bank closures since 2007,” Trepp says in its report.

Trepp developed the T-CAST model in response to the requirements of the Dodd-Frank financial reform legislation. The model was created to forecast bank earnings, capital, and assets based upon the framework and macroeconomic scenarios. Using bank call report data, 12 macroeconomic variables officially released by regulators in November '12 and proprietary algorithms, the T-CAST model produces results for Baseline, Adverse and Severely Adverse Scenarios.