The ABS People Strike Back: Responses To Our Mark-To-Market Posts

Posted by John Carney, Nov 30, 2007, 5:16pm

We’ve now ventured into the dangerous territory of financial accounting three times in the past two days to raise the possibility that the deep discount on asset backed securities obtained by Citadel when it bought E*Trade’s portfolio may amount to a major mark to market event for Wall Street. We knew when we started discussing this that the banks and brokerages would not be eager to make further write-downs to their ABS books. But what we didn’t know was that so many of our readers would be eager to provide the rationalizations for avoiding the mark downs.

We’re told that the bulk of E*Trade’s credit portfolio consisted of prime residential first lien loans. According to one E*Trade document DealBreaker has obtained, E*Trade had assigned a book value of $1.4 billion to mortgages rated AA or higher, $590 million were rated A, and $285 million were Triple B. With the secondary market for mortgages largely opaque—to the extent it exists at all these days—we thought it was rather obvious that a large public transaction in these securities should at least cause accountants to take a second look at their supposedly marked-to-market portfolios.

The main objection to viewing this as a mark-to-market event seems to be that because Citadel bought more than just asset backed securities and because the securities they did buy were a heterogeneous lot, it’s difficult if not impossible to derive a market price for the individual securities. This strikes us as a daring dodge but we’re trying to keep an open mind about this. Please feel free to explain in the comments below why further discounts of ABS books are not called for.

0Recommend this! 10Comments
Posted in: ABS

Comments

Posted by anom, Nov 30, 2007 5:36PM

duh....
timing - it is bonus season! Definately the time of year when wall street talks up its own book.
Further discounts are absolutely necessary but NOT before jan 2008.
look at the abx index, what citadel just paid for these "assests" and the fact that goldman is still short the abx (meaning that they think that some of the abs securties that are trading at 25 cents on the dollar are still overvalued) and one can only conclude "further writedowns". The question is not if but when and the magnitude of the additional discounts.

Posted by Brian, Nov 30, 2007 5:55PM

They did make an investment in E*Trade at the same time that they bought the ABS portfolio. However, Blackrock also invested the senior debt and stock portion of the deal, but not in the ABS transaction, so it would appear that the terms of the capital infusion into E*Trade stand on their own and are not effectively a "unit deal" with the ABS purchase.

While Citadel isn't going to disclose what they valued the securities at, with 60 people on the case, you can bet they have a number for each one. (I don't know who audit's Citadel's books, but you can bet that the other audit teams in the financial services practice of that audit firm might get a peak at what values Citadel put on them - it would be a question the auditor's risk management function would ask at the very least) Moreover, since 45% of the portfolio was in prime first lien RMBS rated AA or higher you can assign some ballpark numbers to the various asset categories here and figure out that the haircut on these first lien, AA or better bonds was severe - even if you assign all the $800 purchase price to these securities, you still come up with a 40% haircut.

E*Trade is alleged to have not bought the pick of the litter when it came to RMBS assets, so maybe their book is worst than most. Still, even putting the best face on it, it will be an eye catching transaction when the auditors sit down to do their work.

Posted by Anominous, Nov 30, 2007 6:21PM

Marking-to-market is done on a security by security basis. It is that simple. There is not enough information available from this transaction to mark anything to market.

And even if Citadel disclosed each security and the price paid for it, that would just be one potential input. Companies could argue that this was a distressed sale and that they should use ABX or rating agency based marks.

There is some price discovery here (although Citadel never pays full price for anything) and everyone knows there will be gobs of additional writedowns, but it has no practical effect on what I have no doubt will be brief and wholly amicable negotiations between the Street and its auditors.

If you want to think in doomsday terms, remember that a lot of the auditors on the other side of the table used to work for Arthur Andersen. Think they're going to be loosey-goosey this time around?

Posted by servicing released, Nov 30, 2007 9:15PM

Hmm.... the ABS portfolio was seperately negotiated and disclosed because Citadel thought their price a) above fair value b) below fair value? The securities transfer immediately avaliable for trading because Citadel thougth the negotiated portfolio price a) rich b) well below realizable value?

I have my opinion, but obviously somebody thinks this is real benchmark... This may be a sweetner to a deal that E*trade would have beenhappy to accept anyway.

Posted by the dude, Dec 01, 2007 1:05AM

I don't question that banks will take additional write downs...what I do question is whether this deal provides any reliable data. ETrade was screwed, they knew it, the street knew it and CIG knew it...if CIG paid .27 then the book is worth at least double that.

Posted by , Dec 01, 2007 1:55PM

Citadel's auditor is PwC. They have a mandatory 5 year rotation with their auditors. D&T did the audit up until 2 years ago.

Posted by , Dec 03, 2007 9:58AM

"if CIG paid .27 then the book is worth at least double that"...

If ANYBODY paid x for ANYTHING, its worth is EXACTLY x. No rationalizing here.

Take the hit now. Use the tax losses and move on (think LEN and its massive tracts of land). Quantify the stink.

Posted by inIT4the$, Dec 03, 2007 10:21AM

I wanna talk about mark-to-model for a sec. There is a market that has been doing this for years without any issues. The problem arises from marking your own stuff to your own model instead of independent pricing (modeling). Anyone wanna take a guess?

Anominous(??)

Given the market above and it's continuing ability to function I would say from this small, relative to the overall size of the ABS market, there certainly is enough information to extrapolate pricing in other similar securities.

Posted by Shaq, Dec 03, 2007 11:51AM

"Companies could argue that this was a distressed sale and that they should use ABX or rating agency based marks."


LOL Yeah, I'd like to see them use ABX-based marks. Many of those would be even WORSE.

Posted by Auto, Dec 03, 2007 6:26PM

No way can you argue that folks shouldn't have to mark their assets down to the price Citadel paid because Citadel bought a distressed asset doesn 't work.

Citadel's purchase was effectively the only one in an otherwise dead market. It just set the market price. Everyone else's prices are just wishful thinking.

Remember all that price discovery stuff from Econ 101?

Post Your Comment