Sunday, August 23, 2009

What FDIC Auctions and Loss Sharing Deals Don’t Tell Us

Given all of the past and apparently forthcoming FDIC action in terms of asset sales and loss sharing agreements in regards to banks put into receivership, it occurs to me that it is not particularly worthwhile to scrutinize the minutia of each deal. Specifically, I think we need to be careful to not put too much emphasis on the value of assets implied by these sales and agreements. This is because there are a number of factors that I believe could lead the discounts to be larger than they would be in an arm’s length transaction. I will discuss two such potentially-skewing dynamics in the following paragraphs.

First, investors have to remember that FDIC asset sales are basically distressed sales. While there may be plenty of bidders, they are all bargain hunters (a nice way of saying vultures) looking to purchase assets with a large margin of safety included in the price. Let me use an example to illustrate this point. Check out this chart provided by Tyler Durden of Zero Hedge:

The loans listed above are commercial real estate loans that the FDIC picked up through its receivership actions over the past year. These are often loans that other banks didn’t want to take on even if they purchased other assets or assumed a failed bank’s deposits. The FDIC is then forced to sell them to private investors and other banks at discounts to book value. Now I don’t know if these loans had already been marked down but the average discount for the above group of loans as sold was a shade over 30%. The majority of these loans were performing, which means that in theory the borrowers were still paying and accruing interest. Accordingly, a 30% haircut on top of any write downs that had been previously taken is a pretty good sized discount. Without knowing anything about these loans it is impossible to say whether this valuation was appropriate or not. But given that this was an FDIC auction it is logical to assume the sale was more like the sale of a home in foreclosure than an arm’s length transaction between a homeowner and a willing buyer.

My point is that we can’t glean too much from this data without a significant amount of additional granularity on the individual loans. Although I believe commercial real estate loans will be increasingly stressed, I don’t want to automatically assume that70 cents on the dollar is the current fair value of a portfolio of mostly performing CRE loans. If that is the case, my guess is that there are a lot of banks carrying these loans at unjustifiably high levels on their balance sheets. And if they are ever forced to mark them down to that level then we will see more banks fall under the minimum capital requirement and potentially be taken over by the FDIC.

Right now according to FASB rules, banks are allowed to carry loans classified as held to maturity or held for investment at historical book value and only have to mark them down if they see a real chance of impairment or move them to an available for sale portfolio. We know that the FASB is considering proposing a change to the current rule that would force banks to mark all of their loans using fair value accounting. Accordingly, while I am anticipating many more bank seizures as the true value of assets eventually comes to light, I think it is unwise to equate the price paid in these auctions to legitimate mark to market values. I certainly am an opponent of the banks being allowed to carry loans at an inflated value even though they would fetch far less on the open market. But I also assume that there is often a value in between historical cost and the price a loan receives in an FDIC auction that better reflects fair value.

Next, let me discuss the recent transaction involving BB&T, Colonial Bank, and the FDIC. As many of you know, Colonial was sold to BB&T after it was taken in receivership and the FDIC agreed to cover losses on certain assets in order to get the deal done. This type of loss sharing agreement is unfortunately not uncommon and was probably the only way the FDIC could find a willing buyer. Associated with this transaction, BB&T included the value it is placing on the assets for which the FDIC will share losses. The chart below details the type of loan as well as the related haircut.

As the data indicates, BB&T is marking down this loan book by an average of 37%. Keep in mind that it is very likely that Colonial (given the distressed condition of its portfolio) had already written these loans down to some degree. Even more troubling is the fact that BB&T decided to mark the construction loan book down by another 67%. Based on data as of March 31st (Colonial did not file a 10-Q for the period ending June 30th), Colonial had $26.44B in assets and $25.16B in liabilities on its balance sheet. According to the press release, BB&T acquired $21.8B of those assets, with $14.3B being covered by the FDIC. Just for fun, let’s see what that 37% mark implies about the bank’s capital condition prior to the sale. Using the data as of March 31st, if you mark those $14.3B in assets down 37% you get about $9B remaining. Then, if you subtract that $5.3B haircut from the $26.44B asset base, you get negative equity (assuming the same number of liabilities) of $4B. Not what I would call well-capitalized.

But, in reality, what does that 37% reduction in asset value decided on by BB&T mean about the true value of those assets? Karl Denninger believes it means that the FDIC waited way too long to act to close a bank that was clearly not meeting its minimum capital requirements:

This is an absolute outrage. How in the hell does a bank get to the point where its construction loans have a real markdown of 67% from par (!), its commercial property loans have a write-down of more than 30% from par, home equity is impaired by 21% and other mortgage loans are impaired by 18%, or 37% on a blended basis, radically exceeding the bank's capitalization, and yet this institution was not seized MONTHS AGO.

Remember, according to The FDIC, Colonial did not (yet) have a negative Tier Capital Ratio, unlike Guaranty and Corus, both of which do! The carnage there has to be at least as bad.

I have been repeatedly asked for hard proof that banks are intentionally misrepresenting asset quality. I have repeatedly pointed to the loss figures from the FDIC, which is hard proof that this has been going on.

Now, I don’t disagree with Karl that the market value of many banks’ loan books is probably well under what the assets are being carried at. I agree that many would either likely have negative equity or would be severely under-capitalized if they were forced to mark to market their loans. It may also be true that the FDIC is waiting too long to take over these banks and thus suffers lower asset values and larger hits to its insurance fund as a result of not acting more quickly to close down these institutions. But where Karl and I differ has to do with the amount of emphasis we put on the additional marks that BB&T took on Colonial’s loan book and the idea that this is proof of intentional misrepresentation.

(Disclaimer: For the following analysis, I used BB&T’s presentation on this matter, the bank’s SEC filings and reports from the major news outlets. However, I also make some assumptions about BB&T’s motivations when it came to negotiating with the FDIC that are somewhat speculative, albeit logical. In all honesty there are a lot of moving pieces and I cannot say I am 100% sure how BB&T is going to account for this transaction when it comes to either an initial write down of the assets or setting up a provision. In any case I think my assessment of the situation is still valid.)

The reason I doubt that BB&T’s valuation is representative of true fair value is due to the bank’s explicit incentives to be as conservative as possible in its marks. If the bank sets expectations low enough the market and investors are apt to be positively surprised by a less dire outcome. It also might make the management team look supremely adept at handling its dodgiest assets. Additionally, by presenting the worst case scenario BB&T was likely able to coerce the FDIC to lower the level in which it assumes 95% of losses. Even though the FDIC will cover 80% of losses from $0-$5B and 95% of losses above $5B, the bank is motivated to paint the most dire picture possible to minimize any additional hits to its capital. The only disincentive to being overly pessimistic has to do with the FDIC sharing in the upside if losses are less than $5B. However, I happen to think that the bank will be happy if the losses turn out to be less than the draconian assumptions imply. In other words, BB&T would probably much rather give some back than be too optimistic initially.

How well covered is the bank? Well, according to the SEC filings, BB&T acquired Colonial’s assets at a $4.5B discount to book value. Then, when you add another $5B buffer from recognizing unexpected hits to its capital, BB&T has created a large margin of safety based on the combination of the original discount and the predicted losses. In other words the entire Colonial portfolio would have to be worth $9.5B less than book value for BB&T to suffer losses it has not already accounted for. But, even a 37% loss of the $14.3B in covered assets (the remaining $7.5B in assets are not covered by the FDIC) is only $5.3B. So, even with the aggressive marks it seems unlikely that BB&T will suffer an unanticipated reduction in capital and since BB&T has already assumed a $5B loss there is no earnings impact unless the losses go above $5B.

Based on BB&T’s incentives, I think we can conclude that just as the 30% discount bidders received in the recent FDIC auction discussed above does not necessarily reflect fair value, neither does the 37% discount that BB&T took on Colonial’s book. Although, it certainly looks like in just about any case Colonial had negative equity. Based on the Q1 data those $14.3B in covered assets would only have needed to be marked down by about 12% to wipe out all of the bank’s shareholder’s equity. Also, the $4.5B discount that BB&T bought Colonial’s assets at was more than 3.5 times Q1 2009 book value. So there is no doubt that Karl is correct in his assessment that the bank needed to be closed or immediately recapitalized.

However, what I don’t think we can do is extrapolate the marks taken on Colonial’s book and apply them to other distressed loan books or use them to make general assumptions about how banks are valuing their loans. Similar to what I said above, it is very likely that there is a number that more accurately reflects fair value and is in between what many loans are being carried at and the value implied by intentionally overly aggressive marks. And therefore, just because a bank is not marking its loans to distressed sale-like values, does not mean it is intentionally misrepresenting the facts. While I am sure most banks lean towards over-valuation, I think the unique economic and financial circumstances should provide banks with a (very) small amount of leeway when determining the appropriate value for their assets. If not, temporary illiquidity in the secondary market for loans could force the FDIC to take over banks that it would be better off allowing to slowly recapitalize. I hate the idea that we have by in large adopted the Japanese mantra of “extend and pretend,” but with confidence in the financial system so tenuous, the last thing we need is unnecessary bank closures.

(Picture courtesy of