Interesting People mailing list archives
Re: A QUERY Credit Default Swap (CDS) question and answer
From: David Farber <dave () farber net>
Date: Sun, 19 Oct 2008 16:40:58 -0400
Begin forwarded message:
From: Tom Gray <tom_gray_grc () yahoo com> Date: October 19, 2008 3:57:16 PM EDT To: ip <ip () v2 listbox com>, dave () farber net, tom_gray_grc () yahoo comSubject: Re: [IP] Re: must read Credit Default Swap (CDS) question and answerProf FarberI hope that you can post this question to your list. The explanation below discusses the pricing of CDOs and their related CDSs. It mentions an unsurprising result that the price of a derivative is related to the hedge for its risk. So the price of a CDS should be related to the risk inherent in the underlying CDO. This is not a result which will surprise many people. So the question is about the pricing of CDSs on the CDOs for mortgages that were sold to people who had no possibility of paying them off. The only way that these mortgages would work was if the housing market rose without interruption forever. This unlikely assumption has been empirically denied.So the question is about how the risk of these CDOs ere estimated. I have read that the people holding the CDOs were indifferent to the risk since they had swapped the risk with the sellers of CDSs.How did these CDS investors assess the risk of these mortgages when they had no relationship with the people paying the mortgages?The deep mathematics that surrounded these instruments would seem depend on this sort of knowledge and be useless without itTom Gray --- On Mon, 10/20/08, David Farber <dave () farber net> wrote: From: David Farber <dave () farber net>Subject: [IP] Re: must read Credit Default Swap (CDS) question and answerTo: "ip" <ip () v2 listbox com> Date: Monday, October 20, 2008, 12:08 AM Begin forwarded message:From: Simon Clift <ssclift () gmail com> Date: October 19, 2008 2:14:53 PM EDT To: dave () farber net Subject: Re: [IP] Credit Default Swap (CDS) question and answer Reply-To: ssclift () gmail com Hi Dave,At the risk of advertising my scum-of-the-earth status as a quant, heregoes an explanation of why CDS's should not be banned or set aside. On Sat, 2008-10-18 at 20:03 -0400, David Farber wrote:From: "David P. Reed" <dpreed () reed com> Date: October 18, 2008 11:39:44 AM EDT Revoking gamblers' lottery tickets? Nice move, unless the gamblers fund your campaigns, as they do in this case.Hedge, please, not lottery ticket. :) Without an instrument which represents a given risk you have two problems. The first is that you have no mechanism for comparing yourestimate of risk against someone else's. No publically visible price leads to a lack of information in the market. Second, when you buy aninstrument like a CDO and think there may be a pricing error (in the pricing of risk), or if the market develops in a nasty way, you needinstruments like CDS's to hedge, or transfer, risk. No hedge leads toinefficiency and costs everyone money. In theory, this comes down to Debaen and Schachermayer's workdemonstrating that (and I'm sure I'm butchering this) the existence of a unique price for a derivative is equivalent to the existence of a hedge for its risk. A derivative, in theory, trades a risk from someone whowants rid of it, to someone who can actively manage it. The cost of management supplies the price.For example, the recent mortatorium on short selling had an immediate,and costly effect on the derivatives market. The first thing traders asked me that morning was "how do we hedge our Puts?", which are contracts that pay when prices drop. Sure enough, the spread betweenbid and ask prices went up and it became more expensive for everyone to hedge their risks. The U.S. government had removed a fundamental marketmechanism and driven up costs. Guess who made money? (Hint: not the purchasers of derivatives.) What would be very useful, IMHO, is a better exchange mechanism for commonly traded "exotic" derivatives like CDS's. The backing of theexchange would ensure both that prices are visible and that payments areguaranteed. Fortunately, the financial world is tending to move this way.and lots of "evidence" that "prediction markets" "always work".There have been a number of studies, actually, indicating that risks implied in the market aren't particulary good (or bad) estimates. Sometimes you can actually extract the pricing model and coefficients that a market participant is using by looking at their quoted prices.It's a career limiting move for anyone entering the math side of US economics profession to be skeptical of the idea that price doesn't reflect information perfectly.The biggest piece of horses**t that goes into any of my models is asingle estimate of future volatility taken from the market. My personalpreference is to attach an estimate of error in valuation inputs then get a price range. That's considerably trickier to do, however, than the usual valuation procedures. On the other hand, since the quoted market price is also the price at which you can offload a risk to someone else, it is inherently the "correct" price. -- SimonArchives __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
------------------------------------------- Archives: https://www.listbox.com/member/archive/247/=now RSS Feed: https://www.listbox.com/member/archive/rss/247/ Powered by Listbox: http://www.listbox.com
Current thread:
- Re: A QUERY Credit Default Swap (CDS) question and answer David Farber (Oct 19)