« Ike Brannon, where is my talk? | Main | The Lehman gift that keeps on giving »
Credit default swap fact of the day
...the CDS [credit default swap] positions of large US banks during 2001–06 grew at an average compounding annual rate of over 80%.
That's from a very good paper by Darrell Duffie. There is more:
Of all 5,700 banks reporting to the US Federal Reserve System, however, only about 40 showed CDS trading activity and three banks – JP Morgan Chase, Citigroup and Bank of America – accounted for most of that activity.
The net transfer of credit risk away from banks is estimated to account for 30 percent of the market. Furthermore a bank may go short on the credit risk of a company it is lending to. A CDS is then a substitute for selling or securitizing the loan. If you think securitization is overdone, CDS has this benefit namely that it is a potential substitute.
A bank also can short the credit risk of a company by dealing in its bonds and other securities. But these other security markets are regulated and replete with restrictions on short sales and the like. The CDS markets don't have comparable restrictions. You can think of the CDS market as, in part, an attempt to circumvent regulations and trading costs in other securities markets.
Here is the single best paper on CDS that I know. Enjoy.
Posted by Tyler Cowen on September 23, 2008 at 05:08 AM in Economics | Permalink
Comments
three banks – JP Morgan Chase, Citigroup and Bank of America – accounted for most of that activity.
All of these institutions of course weathered the storm seemingly fine, so what is the significance of their CDS activity? If the Government had allowed AIG to fail, would the workout of its CDS positions have unraveled these banks as well?
Posted by: Tom T. at Sep 23, 2008 6:45:59 AM
CDS are Teh Pure Evil(TM).
You can not insure yourself against making bad business decisions or in general against something you have an inpact on. This includes decision on signing contracts with solvent counterparties.
Buying CDS to insure oneself against the bankruptcy of a counterparty MUST lead to a situation where your sales department signs contracts with everyone and their dead cats. Great for their quarterly bonuses, a ticking bomb for the company.
In the end you only substitute multiple credit risks against a credit risk with a large insurer. If everyone is doing this, and the CDSs are a large part of the insurers business, then IT HAS TO BOMB.
There was a Dilbert cartoon about this...
Posted by: Oskar Shapley at Sep 23, 2008 7:32:37 AM
"most" for the top 3 means they "control" 92% of all derivatives and 97% of credit derivatives. In both i.r. derivatives and credit derivatives, the big 3 are basically flat i.e net exposure is negligible in relation to notional value.
Posted by: jck at Sep 23, 2008 7:48:45 AM
It's true that the limitations of securities markets are one part of the need filled by CDS's, as you mention. But the overwhelming function of CDS's is that they separate credit risk from all the other risks associated with bonds, for instance. Also, CDS's have similar trading costs, and have counterparty risk, a risk not shared by the securities. I guess I'm saying that you appear to be overestimating the motivation to avoid regulation, but I don't know how much you mean when you say 'in part.'
Posted by: cb at Sep 23, 2008 9:24:53 AM
It's true that the limitations of securities markets are one part of the need filled by CDS's, as you mention. But the overwhelming function of CDS's is that they separate credit risk from all the other risks associated with bonds, for instance. Also, CDS's have similar trading costs, and have counterparty risk, a risk not shared by the securities. I guess I'm saying that you appear to be overestimating the motivation to avoid regulation, but I don't know how much you mean when you say 'in part.'
Posted by: cb at Sep 23, 2008 9:24:59 AM
It's true that the limitations of securities markets are one part of the need filled by CDS's, as you mention. But the overwhelming function of CDS's is that they separate credit risk from all the other risks associated with bonds, for instance. Also, CDS's have similar trading costs, and have counterparty risk, a risk not shared by the securities. I guess I'm saying that you appear to be overestimating the motivation to avoid regulation, but I don't know how much you mean when you say 'in part.'
Posted by: cb at Sep 23, 2008 9:25:02 AM
It's true that the limitations of securities markets are one part of the need filled by CDS's, as you mention. But the overwhelming function of CDS's is that they separate credit risk from all the other risks associated with bonds, for instance. Also, CDS's have similar trading costs, and have counterparty risk, a risk not shared by the securities. I guess I'm saying that you appear to be overestimating the motivation to avoid regulation, but I don't know how much you mean when you say 'in part.'
Posted by: cb at Sep 23, 2008 9:25:09 AM
A PDF of the paper is available direct from the author at http://www.stanford.edu/~duffie/BIS.pdf.
Posted by: Chris at Sep 23, 2008 10:23:05 AM
If JP Morgan Chase, Citigroup and Bank of America were doing a lot of CDS activity, and are RELATIVELY fine [Citi??] ... what firms were, in general, selling these?
Is that side of the business similarly concentrated in, say AIG and federal bailout players to be named later?
Posted by: ZBicyclist at Sep 23, 2008 11:51:40 AM
CDS is then a substitute for selling or securitizing the loan. If you think securitization is overdone, CDS has this benefit namely that it is a potential substitute.
Hedging a credit risk with a CDS creates the same moral hazard as securitization.
Posted by: Nick at Sep 23, 2008 2:54:22 PM
The market treats regulation like damage and routes around it.
Posted by: Andrea at Sep 23, 2008 4:01:36 PM
http://www.revolutioncreditsolutions.com
In the state of this economy and of the nations credit on the consumer side I beleive we will see a revolution. Property values fall gas prices rise and credit scores drop. So much for the American dream, Maybe we can all go to mexico or India ! Thats were they sent all of the jobs
http://www.revolutioncreditsolutions.com
Posted by: Mike Scornavaccco at Sep 23, 2008 6:36:39 PM
I think cb is more to the point. It's not really about going around regulations; asset swaps achieve credit risk transfer, too. Also, i don't know what the papers say, but a CDS is a bilateral OTC contract - it's more like an insurance contract btwn insured and insurer (with the novelty that the insured does not need to own the asset) than a securitization. Yes, a synthetic CDO uses CDS to tranfer credit risk, but as credit asset cashflows aren't being parsed into waterfall, not sure it's helpful to view them as securitizations per se.
The key traits of a CDS are:
(i) pure exposure to credit risk; i.e., the writer of CDS is like the bond buyer, but the bond buyer is exposed to funding, currency and interest rate risk. But the short CDS is long only the credit risk.
(ii) unfunded insurance, unlike a CLN. Enter counterparty risk
(iii) protection buyer does not need to own reference asset. This gives risk to notional protection many multiples of the original.
Posted by: David Harper at Sep 23, 2008 6:39:21 PM
Partnoy and Skeel do a good job of summarizing the issues. But apart from reducing the incentives for banks to monitor, I find the identified problems with credit default swaps somewhat weak.
With respect to incentives to destroy value, while it is true that credit default swaps could play a roll in this the authors concede that they are not at all necessary, for one could achieve the same result by shorting the underlying bonds.
The opacity issue isn’t at all particular to credit default swaps, but rather applies to any OTC product. For example, why should the buyer of a natural gas swap expect the seller to disclose who they hedged the other side with, if at all? Further, nothing precludes borrowers from including language in their credit facilities demanding disclosure of how creditors hedge the risk. This sounds more like a practice recommendation for borrowers than a required regulation to me.
I don’t put too much stock in the industry self protection argument either. The fact is that many standardized contracts in the energy industry (EEI, NAESB, WSPP) are evolving to be more like the ISDA because the later is widely considered a fair and efficient agreement.
With respect to systemic risk, again I don’t see how this is much different than any other kind of derivative (energy, interest rate, weather, etc.). Just about any highly interconnected market is vulnerable to mistakes made by large players. Heck, this applies to trade too.
Also, I find some of the reform/regulation ideas strange. I find stealing ISDA’s intellectual property bizarre. If you can’t afford to buy the ISDA form you probably shouldn’t be trading derivatives. While I’m all for general credit risk disclosure requirements, forcing company’s to disclose documentation on specific transactions is a terrible idea…why should they have to divulge such proprietary information? That’s like mandating that retailers display the wholesale price and the name of the supplier next to the goods they sell. Lastly, there are already plenty of pricing sources for CDS, so why force the creation of a new one? For example, I can walk over to my Company’s Bloomberg terminal and pull up quotes whenever I want…. and Boomberg terminals are ubiquitous in trading shops.
The termination rights section is also curious. The standard ISDA has symmetrical damages language, meaning that the party enjoying a gain as a result of the termination must reimburse the party enjoying a loss, even if the party enjoying the loss is the defaulting party. So sure, plenty of ISDAs with Lehman were terminated over the last week. But those parties that enjoyed gains by terminating will ultimately be sending Lehman a termination payment. They will not be walking off with windfall gains. So what’s the problem then? The creditors to the estate should still be getting the value of those terminated contracts.
Posted by: Chris Janak at Sep 23, 2008 8:11:42 PM
I had some comments on Partnoy and Skeel too, having just read it. The big one is: when they say that there seems to be some very large fundamental mispricing/inefficiency in the CDS/CDO creation markets, is there any way to interpret this in context that doesn't essentially mean "massive ripoff"? If that's correct, why are they (and you, Tyler) so relaxed and upbeat about the whole thing?
Overall, they make a good case that the standard methods for evaluating the CDS/CDOs are very very blunt (standard 0.3 cross-industry correlations? what the $%@#?) if not completely inadequate. Phrases like "swindle", "fraud" and "grotesque incompetence" kept coming to mind while I read their review of the downside.
Posted by: Geoffrey at Sep 24, 2008 6:54:20 AM
I find stealing ISDA’s intellectual property bizarre.
Claiming intellectual property over a class of contracts is bizarre.
Posted by: Nick at Sep 24, 2008 9:53:03 PM
If you have to buy some maple story mesos, please come to our company.
Posted by: maple story mesos at Jan 1, 2009 9:12:37 PM
