The Financial Times headline reads “Bad news on Lehman CDS” when by most readers’ standards, the content is almost entirely good news. It elaborates on and generally confirms the report that we posted earlier from the DTCC, that the net exposures on Lehman credit default swaps by major protection writers was in fact minor. However, as noted in an earlier post, that does not eliminate the possibility that hedge side of a position might fail (that is, a trade to offset the protection written), since hedge funds were reported to have been protection writers too. Even if their involvement was minor in aggregate, it could turn out to be significant for particular firms.
The bad news is that a further Lehman settlement happens on Monday and it appears the prices will be tweaked somewhat lower.
Note the use of the term settlement in the article is somewhat confusing. The auction prices will be set tomorrow. but the actual payment does not occur then. A reader said that the final day for the payment is October 21. While we have not seen the details in the media, the payments would clearly not take place on the date of the auction.
From the Financial Times (hat tip reader Tim):
The pay-outs on around $400bn of defaulted credit derivatives linked to Lehman Brothers are likely to be higher than anticipated after initial results from auctions to settle these credit default swaps resulted in a lower recovery price.
The initial auction results were settled at 9.75 cents in the dollar, meaning banks and other investors who had agreed to make these payments in the event of Lehman’s default will have to pay out 90.25 cents on the dollar.
The final auction for Lehman credit default swaps (CDS) was due to settle in the New York afternoon. The final prices were likely to be lower than the initial ones because there were a lot more sellers of bonds to settle the auctions than buyers, according to traders.
There have been widespread concerns that the prospect of these payouts – the gross value of which is roughly $360bn – is contributing to banks’ hoarding of cash. As well as Lehman’s default, the steady collapse of banks will require pay-outs on other credit derivatives, such as Washington Mutual and three Icelandic banks.
Eraj Shirvani, chairman of International Swaps and Derivatives Association (ISDA), the industry body that manages the auctions, said these concerns were misplaced. “Sellers of protection mark their positions to market every single day. So those firms have already marked down and provided collateral against their positions. As a result, there should be little or no unanticipated additional cost involved in the settlement of Lehman CDS,” said Mr Shirvani, who is co-head of European credit at Credit Suisse.Net exposures were usually around two per cent of the gross amount, which vastly reduced the potential cashflows. Assuming $360bn of gross exposure, this would translate into $7.2bn if these estimates are correct.
“Despite immoderate claims relating to the magnitude of the Lehman settlement, these are insignificant when put into the context of $5 trillion in payments on foreign exchange transactions that occur each and every day,” he said.
The settlement of the Lehman CDS highlights the plunge in value of the debt of the investment bank, which filed for bankruptcy nearly four weeks ago. With around $130bn of outstanding bonds, holders of this debt have made severe losses. The bonds were trading at around 80 cents on the dollar just days before Lehman collapsed. The value of Lehman Brothers’ near-$130bn of outstanding bonds plunged to a new low of 8.5 cents on the dollar after initial prices set on credit derivatives linked to Lehman Brothers came in lower than anticipated.
27 comments
Comments are closed.
This is all confusing and adds to the general chaos sentiment for Joe6pack, so I wish we could just hurry up and get back to the normal Friday bank failures and writedowns, which are more easily understood and easy to ignore.
Does The SEC give settlement information for these types of auctions and if so, are there examples of what happens with this scale of event?
AIG Scandal Grows: Auditor Quit In Protest About CDS Valuations
http://www.clusterstock.com/2008/10/aig-scandal-grows-auditor-quit-in-protest-about-cds-valuations
For example, here’s what Joseph Cassano, the head of AIG’s credit default swap insurance business in London, allegedly told an internal AIG auditor when explaining why the auditor was being excluded from valuation meetings:
“Because…you would pollute the process.”
The auditor quit shortly thereafter and has since relayed his memory of the conversation to Congress.
yves, i think you are spreading a lot of misinformation about this auction. the final price was already set at 8.625% on friday:
http://www.creditfixings.com/information/affiliations/fixings/auctions/current/lehbro-res.shtml
as the ft article mentions, the additional cash payment this results in depends on how much it differs from the most recent lehman bond market price used to calculate the last day’s cash collateral payment. bonds were trading at 13c before the auction. so the additional cash payment = (13% – 8.625%) x 6bn = 260mm
hardly earth-shattering
Pardon me, but this all makes zero sense. Lehman has $130B outstanding bonds that go from 80 cents to 8.5 in a matter of weeks, and $400B of credit swaps based on them only pays off a net of $6B.
I’m sure I must be missing something, but that seems like an entirely worthless set of transactions. No wonder they went out of business if that is what they were selling.
There is no further CDS auction on Monday, for Lehman or anyone else. Nest is WaMu on 23 October.
Count me in. I’m lost.
If for every seller there is a buyer, its a zero sum game, so by definition the sum exposure of the people on one side must equal the sum exposure of the people on the other.
How does that say anything about the ability of the losing side to pay?
How can it be that 90%+ of every person in the market was almost perfectly hedged?
Yes, you mark to market, but these things were priced to assume far lower severity rates so we’re not out of the woods yet
Everything is fine as long as the transactions are funded. Not sure how many more payouts can continue to be funded in the future. Lookout.
anonymous @ 4.49
here is an example:
deutsche bank sells 100 cds protection on lehman to citibank.
citi sells 100m protection to jp morgan.
jpm sells 100m to db.
total market : 300m
net payment : zero (trades net out)
but there is also a second order effect due to daily cash margining of cds contracts that reduces the cash amount even more. here is an example:
citi buys 100m lehman protection from aig
at time of default, lehman trades at 80c => aig must post 20m cash collateral to citi
from time of default up until day prior to the auction, price falls to 13c => day by day aig must post more and more cash (as lehman price falls), reaching 87m on day prior to auction
at auction, final price for settling all cds contracts is determined as 8.625c => aig must pay citi an extra 4.325m on top of the cash collateral it already handed over previously.
now here is something more interesting: why has the lehman price fallen so much? is it because lehman's assets really are totally worthless (well, this probably makes up a large part of the explanations given that stock markets / bond markets have crashed since lehman's default).
here is a further explanation: the cds contract gives the protection buyer an option known as "cheapest to deliver". this means they can choose the bond with the lowest outstanding price in order to determine the final price used for settlement.
the net protection buyers were the big dealer banks that also happen to be the main market makers of these bonds. there is no limit on short selling bonds. so why not target one of lehman's least liquid bonds (which will trade at a liquidity discount anyway) and short it like crazy in the leadup to the auction?
bena gyerek,
I think you are pretty out of line to charge Yves with “spreading a lot of misinformation” when a Financial Times article had a headline “Bad news on Lehman CDS”! The article was not terribly clear, Yves does not appear to have misrepresented it, and its lead author, Aline van Duyn, generally gets high marks from finance pros (held in higher esteem, from what I can infer, than Gillian Tett).
And Yves highlighted the most important bit of news, the confirmation of the netting.
Your clarification is welcome, but you should direct the complaints at the right party, the FT writers, not Yves. Bloggers do not pretend to do original reporting.
In bena’s example, now every day db, citi, and jpm have to be transfering all that margin cash in giant circles every day, on average 61 times around to get back to the beginning. And while all that money is going round in circles, everyone is wonder who the weak link is. No wonder no else has any cash to do real business.
And the guys who designed that are supposed to be financial geniuses? Good riddance.
I agree with the above comments. If the payment is so low, then what is the purpose of CDS? Just to make everyone very interlinked?
I guess what most people are missing is that ****supposedly**** the margin calls have already been done for the case Lehman was debt was set at 13%, so the difference to 8.725% wouldn’t be that bad.
But i really doubt traders would have already set down 300 billion for the payments already, the accounted losses for the crisis haven’t yet reached 600 billion. That would add 300 billion in 1 month. It would have to be accounted for only later but i doubt we won’t see some major hedge fund failures with these margin calls.
Here, directly from the ISDA website is the Lehman protocol timeline:
October 10 → Auction Date
October 15 → Deadline for receipt of Notice of Physical Settlement for trades formed under the Protocol
October 20 → Settlement Date for trades formed under the Protocol
October 21 → Cash Settlement Date for Covered Transactions
There are also some nice graphical timelines for all the protocols here:
http://www.isda.org/companies/ISDAProtocolTimeline.pdf
I am getting lost as well here. If there was 400B CDS written and the sucker went under, how can it be a zero sum game? Someone has to lose while other players have to gain 400B. It is probably not the same entity selling and buying the CDSs. What am I missing?
BTW, Bena Gyerek, udvozlet neked
If margin is being properly done on CDS, why the rush to move to a clearing house?
Perhaps the prime brokers’ credit policies are not as dismal as some say—we’ll soon know.
LEH had some $125 billion in long term debt when it went belly up.
If only $8 billion is the net settlement, then it implies very few, if any, bothered to hedge their bond positions with CDS dealers.
October 12, 2008
As I have said before, anyone who has not read Satyajit Das’s, “Traders, Guns, Money” can’t come close to understanding the complexities of The CDS, CDO, or SDO market. It’s not that you will “understand” anything abut the aforementioned “securities” that is important. What is important is that you will find out that NOBODY understands the complexities of the third, fourth, and fifth layoffs, because they weren’t even understandable then. They were just another “securitized”, hyped up, “product” that the traders laid off to the sales/broker staff in order to meet their bonus max by June/August of each year. And when the bonuses were paid out they all jumped ship, and are now either on the beach of their Caribbean Islands, or working for Paulson, the Wall Street trader equivalent of a “bread line”.
I haven’t ever predicted anything on my previous posts, but here I go: DOW 6,000 by close of Monday Oct. 13, 2008. We shall see.
Earl L. Crockett
Santa Cruz, CA
PS To Yves: My info says that the PAYOUT will be 8.75% of the face value of the mortgage where in turn the holder hands over the security to the insurer. No?
Famed mathematician and hedge fund manager Nassim Taleb, who has written textbooks on dynamic hedging for John Wiley & Sons and was called by recent Nobel Laureate Daniel Kahneman as one of the world's top intellectuals, squares off with Ken Rogoff in the following video, as hediscusses the fallacy of pretending Gaussian distributions in financial modeling is a competent strategy over time:
http://www.youtube.com/watch?v=ABXPICWjFIo
Here is a short blurb on Taleb:
http://en.wikipedia.org/wiki/Nassim_Taleb
Matt Dubuque
fred55 et al,
Thank you for the frank and substantive discussion about this complex subject. I pretend to no special knowledge regarding the markets, just a keen curiosity, and a desire to learn, but I must say, that if no one understands these instruments, or can explain how they work, then it’s only logical that they be eliminated immediately from the financial landscape, pronot. Can anyone tell me whether these “financial weapons of mass destruction” are still being written, by whom, and why aren’t these people regulated?
Thanks!
It’s been reported that Morgan Stanley needs 60 billion in cash to stay in the game this week. 2 things come to mind.
1.) I bet Mitsu makes the terms so unreasonable that the deal breaks.
2.) The swaps may net out at 6 billion but any weak link in the chain causes massive damage to confidence and triggers new swaps.
Any guesses on who the weak link is?
I’ll give ya a hint. Re-read the post!
When I was studying this stuff in the 70’s, they (Robert Merton) used the Central Limit Theorem to estimate probabilities in the tails, which is exactly where the CLT doesn’t work. They are probably still doing the same thing, I don’t know.
I wouldn’t call this a “confirmation” because the source is the same organization (ISDA) for the earlier report.
However, looking at the quote carefully suggests an interpretation of the 6 billion which makes sense: it’s not the total net; it’s the total net *which isn’t already reserved*. So all the net transfers for settling at 13% (the market price before the auction) are going to be real numbers but excluded from the 6 billion). Since Lehman debt dropped about 5% (20 billion) that suggests the “average” Lehman debt goes through 3 transactions rather than the 60-something we’ve been estimating.
3 transactions is not wildly different from the 2 you’d expect from the official policy of market makers like JPB to offset all net trades. So that would be consistent with a reasonable market. The story would be, then:
Roughly 100 billion in net transfers, already reserved and accounted for.
6 billion in new net transfers based on the settlement coming in low.
If these numbers are typical and – very importantly – the payors are generally creditworthy and have indeed reserved for their losses – then the cascading cross-default risk is probably manageable and we are likely to make it out of this with a functioning financial system. It doesn’t rule out an old-fashioned panic, but it does indicate that solvent insitutions can survive a panic rather than being sucked into a giant bankruptcy tangle.
If the other 100 b in transfers has been reserved or accounted for, does that mean that we see ~100b + 6b unreserved change hands on the payment date? Is that when the musical chairs of margin calls finally stops, and we see who really nets out? I too, am confused as to the outcome of this. Has the CDS risk been wildly overblown, or is this just an aberration?
kfunck1: If my interpretation is right, we’ll see 100b + 6b change hand during settlement on the 21st and yes, that’s the end of the Lehman CDS business.
I wouldn’t say the CDS risk has been wildly overblown. This is a huge sum of money, and we’re not yet sure everybody will be able to pay. It does indicate the situation is not as bad as many feared, but people were most upfront with the fact that we really didn’t know.
FairEconomist, thanks for your reply. Does the ISDA publish the results of the settlements? In other words, will we find out who owed who what, and if they indeed paid, or will we just all of a sudden see Bank X and Hedge Fund Y go boom?
So FairEconomist, thanks for the concise summary there. If, as you say payors can, then we have something like a 5% default rate for swaps if I’m following you, which would suggest that the swaps were in fact properly designed. We may hope as much. Our problems are sufficient without a derivative fry up.
Something I would stress generally, which is hard to get in context given _last_ week, is that a significant element in a panic is just that, panic rather than substantive losses. We have real asset losses in the financial system which will become more severe, and equities will come down because their prices do not reflect the economic context or their earnings. But the dysfunction of the credit markets has a significant factor of, well, _panic_ because transactions are not happening to give a context for accurate assessment. This is why the subterfuges of the public authorities to conceal solvency problems and suspend mark-to-market actions are so harmful: they promote actual panic. When assets need to come down, public authorities need to accept this and stabilize the build down, rather than oppose unwinds and blow the stress sideways. Or that’s my view.
I’m not in a condemnatory mood with regard to public financial authorities, but they need a wider perspective of their responsibilities than they have shown over the last five quarters. May we reach that day without having to prise to great a volume of systemic wreckage from off our persons.
Fred55, FairEconomist, et al: Thanks for the additional work; I’d thought I’d been left hanging with my questions at the end of the earlier post “DTCC Claims …”