Stressing The System And The TARP Repayment Race

The administration's draconian enforcement of bonus cuts for "distressed banks" has made Lloyd Blankfein and a slew of Wall Street executives hold their breath until they get the details of Geithner's stress test in order to immediately repay their portion of TARP (although it is unclear if Warren Buffett will dump another billion or so into Goldman at this point). So what will be the likely metrics for the stress test and which banks will be able to pass the test, in order to be on their merry way to TARP repayment?

The rumored stress tests that may be utilized include:

i) tangible common versus assets

ii) tangible common versus risk weighted assets

iii) Tier 1 capital versus weighted assets

The complications arise when one considers the follwing facts:

• Flow of credit to the economy is flat, as bank balance sheets are not growing and new loan originations are down 50%

• Tier 1 capital is at all time highs while tangible common equity (TCE) is at all time lows (gating factor for balance sheet flexibility)

• Banks that can not pass the stress test and can not raise private capital will receive government preferred stock that can convert to common equity to preserve lending in a worse than expected economic envionrment.

• As the testing will initially apply to banks with more than $100 billion in assets, it is likely to extend to all banks that have TARP funding eventually. The initial cutoff was only established to keep the regulatory intervention manageable as the > $100 billion banks hold three quarters of the banking system's assets.

What do the probable stress tests imply?

Three tests are performed, one of which will likely be the version implemented by the administration.

i) Tangible common plus reserves plus pre-provision earnings less cumulative losses as a percentage of assets;

ii) The same as i but using risk weighted assets (RWA) as a denominator;

iii) The same as ii but using Tier 1 capital in the numerator;

The chart below presents the banking universe against each of the test scenarios and compares the ranking versus Q4 Non Performing Assets (NPA) and current P/Tangible Book.

Cumulative loss assumptions are: 25% for construction, 23% for credit cards (9% losses for 2.5 years), 14% for home equity, 12% for consumer excluding credit cards, 9% for first lien mortgages, 6.5% for commercial mortgages, 5.5% for commercial (C&I), 4% for all other loans and 1% for undrawn commitments. Peak unemployment is assumed at 9.5% and peak to trough housing price decline is at 40%.

Results: Banks which test better on average are trust banks such as a Bank of New York and Northern Trust, JPM and PNC due to big marks on acquired books (firesale purchases), and First Horizon, Comerica, City National and KeyCorp due to high tangible common. Banks testing poorly and have high NPAs include Citi, Zions Bancorp, SunTrust, Fifth Third and Huntington Bancshares; also U.S. Bancorp and BB&T do poorly on stress tests however that may be due to NPAs being below industry averages.

Instead of presuming what the tests will look like, it is possible to do a passive risk assesment in the form of a Texas ratio, or simply the ratio of NPA (loans and real estate) to TCE and Loan Loss Reserves.

As the result charts indicate the correlation between the Texas ratio underperformers and the at risk banks per the potential stress test, is very high. Worst scoring (highest risk) on the test are HBAN, C, STI, FITB and MI, while CMA, PNC, JPM, WFC and KEY score as lowest risk.

So what does all this mean for odds of which bank will be the first to be bathing in champagne? Most likely candidates are BK, MS and JPM, as these companies have high Tier 1 ex-TARP ratios (assuming 8% as a threshold). Other potential banks that could crawl out of the bonus grave include STT, NTRS, BBT and CYN. Furthermore, funding could be a consideration, but as bank are growing deposits twice as fast as loans since Lehman and compliments of $150 billion in TLGP, funding would only be an issue for smaller banks. Banks below the threshold line are very unlikely to pay off TARP for the foreseeable future, meaning Vikram will likely be collecting $1 bonuses for many years to come.

With gratitude to GS for charts/data