Two Minds

[peterbirks]

And so the Senate has rejected the bridging loan to the US motor companies. I'm suffering a bit of an internal conflict on this. A couple of my stocks are (unfairly, I feel), perceived to be dependent on the Big Three struggling on as independent companies. And stocks in general are going to get slaughtered today. I can see the FTSE opening more than three figures down (minus 150 would be my guess), because Wall Street is certainly going to open something like 200 to 250 down.

I mean, in the greand scheme of things, it's hard to give a fuck. The bailout failed for the same reason that the companies are doomed -- a failure on the part of nearly all participants to face up to reality. The United Auto Workers may get the main share of the media blame, but there are no innocent parties here.

So, what will happen now? Well, I guess the rosy scenario is a shotgun marriage between GM and Chrysler, which staves off armageddon for a year or so. The non-rosy scenario is that one of these two simply runs out of cash.

The motor companies (annd their workers, and their pensioners) just don't seem to have worked out that creative accounting can only take you so far when it is fighting negative cashflow. GM is just going to run out of money. I think I've written about the credit default rates for the motor insurers many times in the past three or so years. These would then be triggered. Who are the counterparties? Will they be able to pay?

An unhappy Christmas beckons for some people, I fear...

++++++++++++++

It's my last day for the newsletter until the New Year. Hooray!!!!!!


I can now compile lists of all the home things that need doing.

________

Comments

[jellymillion.livejournal.com]

The thing is, the US carmakers don't have a shitload of brilliant new Toyota-beating models that they'll start shipping on New Year's Day if they can just get a couple of bucks to tide them over. GM had puked $50bn in 3 years, not including this year, which I gather hasn't exactly been a laugh-riot.

Too big to fail? They've already failed. Let 'em die, then we can see what, if anything, rises from the ashes.

Bah humbug. Happy Xmas, Detroit.

CDS

[(Anonymous)]

"I think I've written about the credit default rates for the motor insurers many times in the past three or so years. These would then be triggered. Who are the counterparties? Will they be able to pay?"

Peter, I really respect your perspective on things financial, but this is a canard that apparently won't die. The 55 trillion dollar notional market always tossed about is the corporate CDS market, widely confused with other types of CDS that are far more opaque.

The corporate CDS market is not with unknown counterparties, it is with and among dealers and end users. Exposures are measured on a daily basis and are, with the exception of the highest credit counterparties (which no longer include the dealers themselves, who now also collateralize exposures to each other). Therefore, the money has already effectively been paid. For example, if GM is trading at a 90% default probability and a 30% recovery valuation, for every 100 of contract, something like $27 has already been posted as collateral. If they default, that will jump to $30 before the auction is held, and after the auction is held, the net cash flows may well be zero.

When Lehman CDS settled, the net movement of cash was about $3.5 per 100, reflecting the fact bond recovery had been assumed at about 12 the evening before the auction, which settled at 8.5 (October 10 was one of the worst trading days in memory).

Despite all the other shit going down, the corporate CDS market is one of the few things that actually works. Take Tembec, Fannie, Freddie, Lehman, WaMu, Icelandic Banks, 8 CEs in 2 months, several with huge outstanding notionals of CDS, all of which settled without incident in the midst in the most volatile market in financial history. This is not to say that an auto credit event won't be a bitch and a half from an operational settlement perspective, but look elsewhere for where this will cause the real pain.

To be sure, there if you want to know the CDS that is fucked up, check out the standard CDS on ABS product that allowed creation of synthetic MBS instruments when there wasn't enough actual subprime crap to meet yield driven demand. If you haven't read the Lewis article on this latter area, you must:
http://www.portfolio.com/news-markets/national-news/portfolio/2008/11/11/The-End-of-Wall-Streets-Boom

-sxlvr

Re: CDS

[peterbirks.livejournal.com]

Hi SX

Yes, I really must apologize for this. It was written in stream of consciousness and I shoould have thought about it first.

Although a couple of months ago I was unaware of the technicalities of CDSs, I can't claim that excuse now because, only two days ago, I was "correcting" a Wall Street Journal piece on just this matter. The WSJ had 'revealed" $10bn in "undisclosed" liabilities at AIG, which were just such "pay as you go" credit default swaps. Of course, in the case of AIG, where the obscure CDSs had caused all of the problems, it was a reasonable assumption :-)

And, as you say, the CDS market actually works, and I think that this should be shouted rather more loudly. The problems arose from a small minority of crap, not from a systemic failure in the CDS system.

The only potential worry with the "real" CDS market (if derivatives can be thought of as "real" is if the collateral loses its value. Here I fear I am once again at the mercy of being unaware of the deep technicalities, but I assume that much of the collateral is not in cash, but in "cash equivalents" which are thus working twice for the poster. If the value of the cash equivalent falls in value (can you post another person's potential CDS liability to you as collateral for A CDS that you buy?) then presumably you have to post more dosh to "keep it going". This is in addition to the additional collateral required if the general risk of default rises or the recovery estimate falls.

But, as I say, this is inner-workings stuff that none of us has ever really had to worry about before.

PJ

Re: CDS

[(Anonymous)]

Actually, cash is overwhelmingly used as collateral for derivatives exposures (with Fed funds or a ECB equivalent attracted). Standard agreements also provide for posting of treasuries with a small haircut, but that is rarely used.

More esoteric types of collateral would be more commonly converted into cash via a repo financing agreement--quite rare that someone negotiates the right to post weird collateral under an ISDA. I have seen in it specific EMG circumstances, but that might be, for example, Mexican sovereign debt by a Mexican company, and even there the haircuts will be substantial.

Re: CDS

[peterbirks.livejournal.com]

OK, my only puzzle here then is, if cash is overwhelmingly used as collateral for derivatives exposures, and there's $55trn of them floating about, then either:

(a) there's a lot more cash in the world than I thought, because this would make up a fair chunk of it, or

(b) this cash is, in some way, being used more than once. One obvious way would be that if company X posted cash as collateral with company Y for CDS Bannyan, then company Y would consider it fine to post that as collateral with company Z for CDS Cuthbert. Company Z, thence, posts said cash as collateral for CDS Dilbert, which it has with company X. Rinse and repeat.

Is this a reasonable deduction, or does it not work that way?

PJ

Re: CDS

[(Anonymous)]

More headline bullshit. Notional expressed in papers is gross, so if I do 50 10m trades selling protection, and 50 10m trades buying protection, that shows as 1b notional. I shit you not, it's that stupid.

Of course, the tenors of those trades, as well as the spreads set, may not match, so the true offset may not be zero, but in any event the gross notional is massively misleadingly, all but irrelevant to the actual risk, which is better expressed as the replacement cost if my counterparties all disappear.

sxlvr

Re: CDS

[peterbirks.livejournal.com]

This is all v useful. Yes, "replacement cost if counterparty vanishes" would be reasonable.

Curiously, I had a similar problem with a well-known wanky spread-betting company, where I had two currency positions (short dollar and long dollar) of different maturities that, although not true offsets, were damn near close to it. Wanky spread-betting company counted both against my deposit, thus tying up about three grand of my cash. I pointed out that this was wanky, to which they replied with the equivalent of a Little Britain-esque "the computer says no...".

PJ

[real-aardvark.livejournal.com]

There's a point in Peston's otherwise largely meretricious book where he says "This description of how a CDS works is probably difficult to understand. Against all my instincts as a writer, I think that may be a good thing."

(I say meretricious, but that's liable to be misinterpreted. Peston himself is an excellent and honest writer. It's only the book that sucks.)

There seems to be a general herd opinion out there that "this stuff is difficult to comprehend." Given sufficient transparency, it most certainly is not. In this particular case, my thanks to both of you for actually discussing the underlying mechanisms, rather than throwing several trillion pieces of verbal confetti into the air.

[peterbirks.livejournal.com]

A CDS really isn't complicated at all. In fact it took me longer to understand interest rate swap deals (and I still hesitate to try to explain them to other people).

If you think of a CDS as an insurance contract, it's simplicity itself. I own municipal bonds that are rated triple B. This doesnn't suit my risk portfolio, which requires rather more single A rated or above investments. So, rather than sell the muni bonds and buy some other A-rated bonds, I just "insure" against the bonds defaulting with another company (which has a high credit rating). For a fee, this company promises to pay up if the other company defaults.

Now, since bonds from triple B-rated companies pay more than the equivalents from single A-rated companies (because they are 'riskier'), the 'fee' roughly equates to the difference in the interest rate.

So, effectively, the buyer of the CDS gains extra security for a lower effective interest rate, and the seller of the CDS gains extra income in return for a higher interst rate — without the need for the two companies to sell one bond and buy another.

As systems go, it's fairly cool.

The problem arises when you buy the CDS (or sell it) without having an underlying interest. As is the case with all derivatives, if this gets out of hand, you end up with a very large tail wagging a very small dog, distortions in pricing develop, which in turn distort the fundamental base market, which makes companies that aren't that risky look riskier and companies that aren't that safe look safer.

And let's not head into the problem of what happens when a rating agency changes the underlying rating....