Tuesday, March 31, 2009

PPIP: Maybe you'd like it back in your cell?

The big question surrounding the toxic asset plan is will banks sell? I've put a little pencil to paper here and come up with some actual numbers.

First of all, I expect the Legacy Securities Program will work wonderfully. Sellers will flock to it like Jawas to a stray astromech droid. This program is aimed at securities which have been severely impaired from both a credit and liquidity perspective. CLOs, RMBS, ABS, CMBS, etc. The program should succeed in turning these programs into just credit impaired. In effect, it will separate the red ones with the bad motivators from the blue ones in prime condition. That will be key to an eventual economic recovery. It should foster a healthy new issue market for RMBS, ABS and CMBS (don't know that CLOs can come back), which will help get the velocity of money back to a more normal level.

Obviously having ready buyers able to earn impressive ROEs should improve the value of the underlying assets. Financial institutions have already marked these securities to market, an improvement in the actual value of the instruments ought to result in an improvement of balance sheets. This will particularly benefit financials who invested primarily at the top of the asset-backed capital structure. It will also benefit those that hold more risk in securities (such as brokerages Goldman Sachs and Morgan Stanley and possibly some P&C insurers) and less those that hold risk in loans (such as almost all banks). Even there, much of Goldman and Morgan's risks are tied to the equity markets, not to debt markets. Same goes for life insurance, generally speaking.

That brings us to the with the Treasury's Legacy Loan Program. I expect this to go over like the exotic twi'lek dancer's routine in Jabba's palace. The plan will indeed increase the theoretical price at which banks could sell loans. That's fine, but its short help. Loans haven't been marked to market. Instead, they are held at book value less an allowance for expected loss.

I've done some deep dives on bank residential loan portfolios. Getting detailed data is a challenge, but basically I tried to figure out what percentage of the bank's current portfolio is "challenged." High CLTV, bad geographics, low FICO, etc. You can make a relatively safe assumption that most of the loss reserve is pledged to those kinds of loans. Anyway, I can't find any big banks that are holding, say home equity loans at less than 90% of face. Unless the Legacy Loan Program winds up buying assets at $90 or more, banks won't sell.

So will the PPIF's pay $90? I doubt it. Take home equity loans as an example. Start with the following assumptions:

  • PPIFs get loans at 1mo-LIBOR +50bps. Its hard to say exactly what the cost of funds might be, but worth noting that FDIC paper trades around L+20.
  • 6-1 leverage, which is the max allowed under the program. I think its reasonable that non-delinquent, prime loans would get the max leverage.
  • Assume the loans float at Prime-flat.
  • Assume the loans are 1-2 years old, and will repay over the course of 6 years. To make it easy I'm going to assume equal payments per month.
  • The pool of loans will suffer 10% cumulative losses, all of which occur in the first two years. I won't write down the losses as they occur, simply take away the interest. That's consistent with a hold-to-maturity IRR calculation.
  • Finally, and perhaps most importantly, I'm assuming that the PPIF equity investors are targeting an IRR of 20%.
The result? $82.5.

That price will render it impossible for most banks to sell. For example, based on Bank of America's recent earnings presentation, it has something like $250 billion of prime, non-delinquent home equity loans with 90%+ LTV. I'd call this the kind of stuff that isn't exactly toxic, but selling could improve BAC's risk exposure significantly. Say they effectively have a $90 mark on these. If they sell at $82, they'd suffer an 8% loss versus their capital, or $20 billion. BAC has core equity capital of $48 billion. You do the math.

I don't expect commercial loans to be much better. Now a lot of commercial stuff has large loan loss reserves, and therefore sales would more easily be accretive to capital. But commercial loans are also present an information asymmetry problem. You can put a zillion residential loans into a pool and get some semblance of diversification. You can then look at average stats and get some idea of the make-up of the loans: geo diversification, average FICO, etc. A bank that is selling a commercial loan is telling you they don't want that commercial loan anymore.

This isn't to say the toxic asset plan will have no positive impact, but it is likely to be more indirect than investors are currently hoping. The best chance banks have for decreasing their residential loan portfolios is a revived securitization market, which is the primary aim of the TALF. Banks may be more willing to sell a portion of their home equity loans as a senior security, with the bank retaining a subordinate position. In that case, the bank might retain the upside while still freeing up some capital. In addition, a revived securitization market would give the market confidence that banks have enough liquidity to hold their loan portfolios to maturity.

It will also help banks who have made larger writedowns, especially those that made acquisitions. At the time of acquisition, the bank has to write down the loan to fair market value. In the case of J.P. Morgan's acquisition of WaMu, or Wells Fargo's acquisition of Wachovia, there would be no motivation to under-estimate the FMV decline. Those banks could therefore enjoy improved capital positions from certain sales.

12 comments:

Keith said...

I'm no accountant, but I'd think with a loan on the books you could sell it at your book value minus the amount of loan loss reserves you had to allocate against it and you would still break even, as those reserves would be freed up by the sale. Though even these days is that more than a few bps? Maybe for second-liens or HELOCs the severity is high enough for that to be significant.

massiveF said...

I love your blog. Thanks very much for the great writing an please keep up the good work. I have some questions about your post.

* Do you think the Legacy Securities Program will still create a revived securitization market when the funds can only be used for securities issued prior to 2009?
* Do you think the banks could be accurate when they carry their loans at $90? If they thought they were worth less they probably would have already sold them off in a security. Why do you think a bank is more likely to sell you junk in a commercial loan then a bunch of residential loans?
* How "retail" do you think PPIFs will get? Can I buy a $100 slice of something? do you think this could be a good idea?

Thanks again for the great blog.

Unknown said...

I don't think you've got either the concept of "capital" correctly understood.. nor accounting..

What do you imagine happens to the money that you sell the loan for? your 82 dollars say... Does it go into the ether???

taxes are a bigger issue than the exercise you went through..

Accrued Interest said...

Wriight: If a bank can sell an asset at book value, including the LLR, they'll do it. But if they need to take a large loss, they won't.

Massive: I'll write more posts on this subject soon.

Robert: I imagine that you sell an asset for a loss its a negative event for capital. I don't understand your question. Nor what taxes have to do with it. Are you talking about the AIG bonus BS?

mojakus said...

AI - you're right about the mark vs. the levered bid: no sellers (apart from JPM/WaMu, BofA/CW, WFC/Wach and PNC; these guys happened to buy loans sub-par) will voluntarily sell at a substantial discount because it would kill their capital. But the key operative is "voluntarily": another way to understand the Legacy Loan portion of the PPIP is to think of the FDIC as creating its own demand. There's an upcoming wave of forced bank closures, and the Legacy Loan program is the method the FDIC intends to use to handle the flow.

Accrued Interest said...

Mo:

There is something to that. I saw JPM threw out this idea in a recent research report. Seems reasonable to me.

Anonymous said...

This video explains another problem of Geithner's Plan:

http://vodpod.com/watch/1470235-geithner-plan-the-problem-of-banks-buying-the-assets-from-themselves-22?pod=misstrade

M.G. said...

As I have written elsewhere this plan will fail because we are in a market for lemons with high level of asymmetric information (it’s not only clarity or liquidity). Sellers and buyers of toxic assets will not reach an equilibrium price or discover the “right” price unless somebody cheats. Uncertainty on the cash flow of the toxic assets is still there so it is the lack of confidence in and among the banking counterparts. Moreover there is no money around for this further securitization of toxic…

Credit Cruncher said...

That's a very useful post, thanks, especially on the pricing for the banks. I am finding it less easy to dismiss this legacy loan program as something that will be an obvious, immediate failure, due to the 'bribing' of large money managers with lots of management income and assets under management potential, along with the asset managers' ability to creatively structure their products, as well as some still existing credibility with retail and institutional investors. I'm not sure that the PIMCO's of the world can't mix in some good and bad assets together, stamp it with a diversification stamp, mix in a little 'trust us', and take on enough 'bad assets' to label this a success. It may be a superficial fix, but it would also be hard to know from outside. I think Geitner was very smart to align the interests of these players to his cause and it will be interesting to see how it plays out.


Robert, unless I am mistaken, the answer to your question is that a bank uses the 98 cent cash from the sale price to pay off the 95 cents of debt it used to buy the original asset, then is forced to write down its equity from 5 cents to 2 cents on the loss. I think AI understands it correctly, unless I am missing your point as well...

Credit Cruncher said...

Whoops, or some number that makes my math actually add up. ;)

shivz said...

There is an operational solution, circumventing the initial toxic assets' valuation obstacle, making the banks sell, giving the banks' stockholders a fair chance and, eventually, seizing control of insolvent institutions whenever necessary.

This is how it could be done:

1. Banks should be obliged, under suitable legislation, to sell their toxic assets to the government at their book value (i.e., accepting the banks' own valuation as per their last statement).

2. Instead of paying cash, the government will provide a guarantee on an amount equal to the value of the purchased assets. The terms of the guarantee will include, inter alia, a formula, according to which, ownership will be eventually apportioned, as well as provisions re management.

3. The existing and/or appointed banks' managers will be given (say) three years to sell, as trustees, said toxic assets (possibly, under economic recovery conditions).
The proceeds will be paid to the banks, concomitantly reducing the amount of guarantee. Thus, the higher the price they get, the lower would be the residual amount of guarantee, and vice versa.

4. At the end of the three-year period, or after the completion of the sale (whichever comes first) the ownership issue will be settled according to said formula, the core of which is the size of the residual amount of guarantee – the lower the amount, the higher the prospects of stockholders' salvation (unsold assets will be valued, for that purpose, at zero).

5. Cash transfers from the government to the banks during the period will be made only in cases of liquidity problems (subject to the terms of the guarantee), and will be taken into account in arriving at the final ownership apportionment.

Note: assuming that, as a matter of policy, uninsured creditors are meant to emerge unscathed from this crisis, the whole amount of losses in banks 'too big to fail' is bound, by definition, to be borne by taxpayers and stockholders. This plan, whilst being the fairest possible to stockholders, makes sure that they are wiped out before taxpayers' money is poured in.

Unknown said...

Sad to see the "cure" for overleveraged, undercollateralized balance sheets to be... more leveraged finance.

Direct recapitalization of banks and a longer duration loan-writeoff period would eliminate this problem entirely (government allows losses to be amortized over 10 years, but only for losses to be declared in the next two quarters, and the favorable terms end after that).

Say citi/jpm/bac has 80billion in writedowns coming - allow them to write it down 2 bucks per quarter for the next 10 years. Internal revenue generation will cover those writedowns - but only if they are upfront and honest about it immediatly.

Taxpayer doesnt pay a dime, private investors stop pricing armageddon into the financials and we recover just in time for the current cohort of traders/risk managers/regulatorslegislators who created this mess to quietly pass the torch to the next generation - who has hopefully learned a thing or two from this.