Competing for Customers and Capital

Stress Test Results: AQ Chances vs SCAP Buffers

On Tuesday morning May 5, 2009 I was interviewed by Connell McShane on Fox Business News live about my recent article Stress Testing: What’s Your Bank’s AQ? I spent a lot of time positioning the AQ score and describing how the good banks did. Then Connell asked me this question:

Wells Fargo is, forgive me, on the [Fed’s] naughty list … but Wells Fargo is number two on your [Good Bank] list. Is that right?

My answer to his question was: Yes, the AQ score for Wells Fargo & Company (WFC) was +5. This made it viruatlly certain the company had enough capital in the last quarter of 2008 to absorb future losses if the recession deepened. Connell then asked me to cmpare my AQ scores with the Fed’s stress tests. I said the major difference is that the AQ test is simple and fully transparent.

THE BUFFERED BANK AQ SCORES
Late Thursday afternoon May 7, 2009 the government published its stress tests of the 19 biggest banks. Except for privately held GMAC all of these institutions were included in my Seeking Alpha articles on the AQ index.

So now it’s possible to compare directly my AQ predictions with the SCAP simulations pointing to the nine banks that required a capital buffer to sustain a prolonged worst case scenario.  The following table compares results for the nine "buffered" banks shared by both studies.

The SCAP report concluded that all nine banks required a capital buffer. Based on their AQ score the chances these banks were holding a significant proportion of troubled assets on their balance sheet appear in the last column of this table.

Of the nine “buffered” banks, the AQ predicted eight of them correctly. And six of these predictions were virtually certain with chances greater than 0.96. The other two were pretty good bets: 0.74 for Fifth Third Bancorp (FITB) and 0.81 for Morgan Stanley (MS).

Then there’s WFC: its AQ chance of needing capital was virtually zero. This was not a near miss. Its AQ score went off the chart.

I knew the Wachovia acquisition would have an impact on Wells Fargo, but I had no evidence to support the rumor that WFC was on the Fed’s “nasty” list. In the last three quarters of 2008 Wells Fargo’s AQ score got better and better. Topping out at greater than +5. That’s why its chance of having a significant proportion of those troubled assets on its balance sheet was zero. Sudden swings in equity and debt market performance are tricky, especially in the interval between quarterly financial reports during the volatile markets of late 2008 and early 2009.

UPDATED WELLS FARGO AQ SCORE
As soon as Wells Fargo filed its report with the SEC on Saturday May 9, 2009 I downloaded the statements from I*Metrix. Here’s what happened between the company’s September 2008 and its March 2009 filing in which the Wachovia merger results first were reported:

• WFC’s revenues almost doubled to $24 from $13 billion
• The price of its common stock fell more than 50% to $14 from around $38.
• The number of shares outstanding rose by over 30% from 3.3 to 4.4 billion.

These numbers turned the company around – in the wrong direction. Wells Fargo’s share of market capitalization fell from 12% to 6%. And share of revenues jumped from 8% to 15%.  The difference between revenue share and value share in the set of 92 banks on my list went from a little over +4 to just under -9. After adjusting the raw score for volatility Well Fargo’s AQ plunged from +5 to -4. Investors in the equity and debt markets declared the merger a bad idea. The Fed agreed.

In the first quarter after completing the merger with Wachovia the AQ chances that Wells Fargo had a significant proportion of troubled assets on its balance sheet went from zero to near certainty. This not only explains why my earlier prediction was “wrong,” it also validates my assertion that the AQ is both simple and transparent. It might even produce useful information.

Thank you for visiting. As always your comments are welcome.

~V

p.s. Here’s a link to my Fox Business News interview. It begins around 34 minutes into the hour. 

Stress Testing: What's Your Bank's AQ?

The Federal Reserve will not be releasing the results of its stress tests of the nation's biggest 19 banks until next Thursday, and undoubtedly there is some element of folly in daring to beat the Fed to the punch. After all, according to The New York Times, the tests have been in the works for two months and have involved more than 150 regulators who, in addition to poring through voluminous bank data, have had the benefit of the banks' analyses of how various hypothetical situations will affect their loss rates.
 
Still, one can't help but wonder if there are other tools at hand that can get the job done properly without taking so long. In fact, I introduced such a tool in my last post – a bank’s asset quality [AQ] index. It enlists an engine of analysis that dwarfs even the Fed's 150 regulators -- namely the equity and debt markets. Quite simply, it compares a bank’s share of market capitalization to it share of revenues -- the interest it earns on producing assets plus fees for banking services. This provides a quick insight into how the share of value investors assign to these institutions compares with the share of revenues they earn on assets, which constitutes a direct measure of their investment quality. 
 
When share of market capitalization exceeds share of revenues, it suggests that a bank may be holding the amount of common stock necessary to absorb future losses; when the reverse is true, it suggests that it may not. Doing the calculations doesn’t take two months and 150 regulators, although it does focus on something that the Fed defines as essential to banks' health -- the amount of capital represented by common stock.

CALCULATING YOUR BANK'S AQ
Calculating a bank’s AQ score is a simple two step process:

Step 1 is to decide on a sample of banks: three or more banks over ten or more periods.
 
Step 2 is to sum their total capitalization and revenues and calculate the percentage of the total represented by each bank. Then take the difference and standardize it. Statistics 101 shows us how to do this in units called standard deviations, which are measures of how far a bank’s AQ score is from the mean. This is the bank’s AQ – Asset Quality.
 
Bank AQ scores above the mean are positive and those below the mean are negative. Statistics 101 also teaches us that for any Bank AQ value in this sample greater than +1.7 the chance that has troubled assets is less than 5%, whereas any value lower than -1.7 means that chance is greater than 95%.

THE GOOD BANKS
In terms of the Federal Reserve’s criterion, less than a 5% chance of holding troubled assets means it is highly likely these 40 banks have enough capital in common stocks to absorb future losses as the recession wears on.

For comparison, the April 24, 2009 DealBook story “Judgment Day for Big U.S. Banks” concluded that U.S. Bancorp (USB) and JP Morgan Chase (JPM)  “probably have put the bulk their troubles behind them.” These conclusions are supported by their AQ scores of +4.15 and +1.81 respectively  -- these banks almost certainly have put the bulk of their troubles behind them.

THE MIDDLE BANKS
Based on their AQI the 32 middle banks in this list had a greater the 5% but less then 95% chance of holding troubled assets in their portfolio in the last quarter of 2008.

In terms of the Federal Reserve’s criterion, the banks in this list have an increasing chance of holding troubled assets. This means it is highly likely that those banks near the top of this hold enough capital in common stocks to absorb future losses as the recession wears on. Those near the bottom of this list likely do not. Banks in the middle have around a 50/50 chance.

That DealBook story also concluded that Bank of New York Mellon (BK) and State Street Corporation (STT) probably have put the bulk of their troubles behind them. Based on their AQ both of them are on the cusp between the good and middle banks with scores of +1.66 and +1.36 respectively giving them just a 0.05 and 0.09 chance of having future troubles as the recession wears on.

THE UGLY BANKS
Based on each bank AQ the 20 institutions in the following list have a greater than 95% chance of holding a significant proportion of troubled assets on their balance sheet.

DealBook also concluded that a number of regional banks are “strong candidates for needing additional money.” The story fingered Sun Trust Banks (STI) with an AQ of -2.14 and Regions Financial Corp (RF) with an AQ of -2.93. These AQ scores are associated with a 0.98 and .099 chance of needing additional capital.

THE SAMPLE
In the American Bankers’ list of the top 150 institutions with the most deposits, 112 were public companies. Twenty of these failed to report in one or more of the 32 quarters ending with the last quarter of 2008. The final sample of 92 “Bank and Thrift Holding Companies with the Most Deposits Q3-2008” was used in this study.

CONCLUSIONS
One might call this method of assessing a bank’s health no more than taking its temperature. But, even if temperature doesn't provide complete information of the health of a bank it does sort out the ones in need triage.

Thanks for visiting. As always your comments are welcome.

~V