Chart of the Day: Common capital vs TCE

By Felix Salmon
May 9, 2009
This chart comes from today's WSJ , and shows the big difference between tier-1 common capital, which is the criterion that Treasury ended up using in its stress tests, and tangible common equity, which is the criterion everybody thought Treasury was going to use in its stress tests. ... But judging by where Wells is right now, it's 90 basis points short of the 4% common capital ratio, but only 10 basis points short of what everybody thought was going to be used: a 3% TCE ratio. " data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

stress.jpg

This chart comes from today’s WSJ, and shows the big difference between tier-1 common capital, which is the criterion that Treasury ended up using in its stress tests, and tangible common equity, which is the criterion everybody thought Treasury was going to use in its stress tests. And you can see why Wells Fargo, in particular, was livid about the switcheroo:

Negotiations with Wells Fargo, where Chairman Richard Kovacevich had publicly derided the stress tests as “asinine,” were particularly heated, according to people familiar with the matter. Government officials worried San Francisco-based Wells might file a lawsuit contesting the Fed’s findings.

Remember that the numbers in the chart are as of year-end 2008, not the year-end 2010 figures used in the stress tests. But judging by where Wells is right now, it’s 90 basis points short of the 4% common capital ratio, but only 10 basis points short of what everybody thought was going to be used: a 3% TCE ratio.

Banks like Bank of America and PNC, however, clearly benefitted from the change: they’re both short on TCE, but have much more than 4% tier 1 common capital.

Why did Treasury switch from TCE to the even-more-obscure common capital metric? Quite possibly to help Bank of America and Citigroup get the amount of capital they needed to raise down to a number within the realms of possibility. After all, these tests were designed so that they couldn’t be flunked. And that might have seemed a real possibility back when Treasury was still using TCE.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
Comments
One comment so far

The stress tests were never really about testing stress but mostly about putting confidence back in the market. From where I’m sitting it looks like it worked. There is still a risk that some unknown unknown might pop up and deflate the new confidence bubble, but for now it looks a bit rosier. What we all have to bear in mind is that black swan events may have low probability, but that doesn’t mean they can’t happen one immediately after the other. While banks remain ready to repeat the errors of the past, the problem is still there. I can’t quite see anyone really wanting to turn off the Wall Street money spigot, so risky lending is probably simply waiting for the encore.

Posted by Nic Fulton | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/