The role of derivatives in the 2008 crisis

“More generally, the 2008 credit crunch was never related to worries over traded derivatives; it was — like all credit crunches — related to much more general worries over bank solvency and the quality of banks’ balance sheets.”  Felix Salmon

The claim that the 2008 crises were not related to worries over traded derivatives is simply wrong.  And I write this post to present some of the considerable evidence against this claim.  (That said, I agree with Felix Salmon’s critique of this article that the Greek default is not a good example of CDS-generated systemic risk.)

First, William Dudley the President of the New York Federal Reserve Bank has stated publicly that:  “The novation of OTC derivatives was an important factor behind the liquidity crises at both Bear Stearns and Lehman Brothers.”

Novation takes place when, for example, a hedge fund decides that it doesn’t want to face Bear Stearns as a counterparty and therefore transfers the contract from Bear Stearns in order to face JP Morgan Chase.  In the process any collateral the hedge fund has posted to Bear Stearns must be transferred to JP Morgan Chase.

Second, the OCC derivatives reports (that include derivatives held by regulated commercial banks, but not by the investment banks) indicate that credit derivatives experienced much greater changes in values than other derivatives over the course of 2008.  Specifically credit derivatives comprised 20% of the fair value (summing positive fair value with negative fair value) of derivative portfolios in the first quarter of 2008.  (See Table 6 here.)    And by the third quarter had jumped to 32% of the fair value.  (here)  Thus, it reasonable to conclude that a minimum of 20% to 30% of the novations that played an important role in the collapse of both Bear and Lehman were due to credit derivatives.

It is entirely possible that significantly more of these novations were due to credit derivatives.  Since the aggregate fair value of bank derivative portfolios actually fell from Q1 to Q3 2008 by 6% or $250 billion, the increase in fair value of credit derivatives reflected a move from approximately $1 trillion to $1.3 trillion.  In other words, as one might predict, the value of credit derivatives moved much more dramatically during the “credit crunch” than the value of other derivatives.  Under the circumstances of these dramatic price changes, it seems safe to assume OTC derivative related collateral calls between Q1 and Q3 2008 were more likely to involve credit derivatives than other derivatives.  And I would posit that such collateral calls can trigger demands for novation.  Attibuting 30% of the novations in the credit crises of 2008 to credit derivatives is therefore probably a low estimate.

[Update 3-2-10:  The ISDA Margin Survey 2009, Table 4.2, indicates that from 2007 through 2009 66% of credit derivative exposures were collateralized.  Thus when the fair value of credit derivatives held by regulated banks increased by $300 billion from March 31, 2008 to September 30, 2008, it’s fair to assume that this had the direct result of increasing the demand for collateral over this period by $200 billion.  And this number excludes the increased demand in collateral created by the investment banks’ credit derivatives portfolios.]

Finally, we have the fact that New York Fed jumped on the CDS market after the Bear Stearns bailout – and didn’t allow Lehman to fail until after “centralized settlement among major dealers” for credit derivatives was implemented.  Previous industry commitments with respect to credit derivatives were focused on back office infrastructure issues and had a leisurely timeframe.  (This September 2006 press release indicates that significant advances for the industry included an end to novation without consent, an 80% reduction in the number of confirmations outstanding for more than 30 days, and a significant increase in the confirmation of trades on an electronic platform.)

In March 2008, by contrast, the industry agreed to automated novations processing by the end of 2008, and full implementation of centralized settlement among major dealers by September 2008.  The latter commitment was confirmed by the dealers in July 2008.  (page 3 here.)  In other words, the regulators moved from jawboning to a demand for commitments from the credit derivatives dealers for major changes in the clearing process that included very close deadlines, due to concerns over “the resiliency of the OTC derivatives market.”

While Felix Salmon may argue that this sudden seriousness about reform of the credit derivatives market was pure coincidence, most would conclude that it reflected regulators’ concern about the market after the failure of Bear Stearns in March 2008.

Overall, the claim that “2008 credit crunch was never related to worries over traded derivatives” is contradicted by the facts.  Insiders have acknowledged that novation of OTC derivatives played an important role in the crisis.  And I think Felix Salmon would have difficulty finding a member of the industry who is willing to assure him that the credit derivatives market would have handled the failure of Lehman effectively, if it had still been working with the market infrastructure of March 2008. Given this situation, the evidence points pretty clearly to likelihood fact that the CDS market was one of the reasons the Federal Reserve was unwilling to let Bear Stearns fail the way it let Lehman fail.

An additional note.  Felix Salmon writes, comparing bilateral clearing to a clearinghouse:

“If a bank has written so much credit protection that it becomes insolvent, then there’s significant systemic counterparty risk regardless of how its derivatives trades are cleared.”

Huh? The whole point is that the bank default to the clearinghouse is born jointly by the banks that guarantee the clearinghouse, instead of being born by individual counterparties – who are naturally more likely to be bankrupted by the bank’s failure to pay, then if they share the losses and thus divide them by 3 or 5 or some such number.

The point of a clearinghouse is not to prevent bad banks from facing a credit-crunch.  Obviously from an efficient markets point of view the sooner that happens, the better.  The point is that clearinghouse should help prevent the credit crunch/failure of a bad bank from taking down other banks that could potentially be made insolvent by excessive exposure to the bad bank — and thus to prevent the bad bank from setting off the chain of failures that defines a systemic crisis.  Of course, if the banks are willing to rely on collateral and choose not to limit credit to bad banks, then we’re in big trouble in either system.


9 thoughts on “The role of derivatives in the 2008 crisis”

  1. “most would conclude that it reflected regulators’ concern…”

    Not that I am nec. a huge cheerleader of the stability/innocuity of CDS or any other derivatives. But these two things (a) what actually caused the credit crunch and (b) what regulators, later, worried about and tried to ‘fix’ – do not have to be the same or even related to each other.

    1. I understand that one can claim that the problem wasn’t the derivatives themselves (or the repos), because they were just the tools by which the credit crunch was realized. That is, in some philosophical sense the “cause” of the crisis was the losses on subprime mortgages, CDOs, etc, whereas the repos and the derivatives were just the evidence of some deeper phenomenon.

      But I’m not really convinced that it makes sense to say that the reason that the house of cards (i.e. the collateralized derivatives/repo system) fell down was the fact that you opened the door and let in a gust of wind (i.e. the weak mega-bank that was guaranteed to eventually test the system).

      And, of course, the issue I had to leave out of the post above for lack of data is the question of what fraction of the losses that threw the solvency of our financial institutions into doubt were synthetic. I’ve never seen a good breakdown of how exposure to synthetic RMBS and synthetic CDOs (including those nested in other structures) created losses as compared to losses from purely cash asset securitizations. (Were there any of latter originated in the later years of the crisis?). So when you tell me that the losses were caused by RMBS and CDOs that sounds to me as if a good fraction of the losses were caused by losses on synthetics. But, of course, I’d love to see solid data on this issue.

      1. I was saying something different & more basic: let’s not give regulators credit for omniscience when it comes to accurately diagnosing ‘what caused a crisis’. The fact that they ban X while saying ‘ prevent Y’ is not evidence to me that X had anything to do with Y.

        To your point here, even if CDOs (or whatever) blew the house of cards down – I don’t think it makes sense to blame such a thing for ‘causing the crisis’ without at least acknowledging it, almost by definition, must have also ’caused the bubble’ – which makes any account of blame and damages at least trickier and more complex to unravel. Is a world where housing prices have been stagnant since 198X so obviously better than our current situation?

        Re: cash/synthetic – I do not view that as an important distinction, but I think we’ve talked about that before. If I had to guess I’d guess that the [synthetic]/[cash] proportion only increased throughout the bubble, that 2002 vintages were healthier and less bubbl-icious than 2007 vintages, and so (therefore) if only for that reason alone a census of ‘CDOs that caused losses’ would be disproportionately synthetic. But of course I see all that as a symptom not a cause of anything. There is nothing special about ‘synthetic’ CDOs that makes their losses hurt any less or more than ‘cash’. A loss is a loss.


      2. “The fact that they ban X while saying ‘ prevent Y’ is not evidence to me that X had anything to do with Y.”

        But it is clearly evidence of the regulators thought process and the likelihood that concerns over CDS markets (whether or not they were correct) probably played into the decision to “rescue” Bear Stearns from failure with cash from the Federal Reserve. And the whole discussion in this thread is whether or not there were “worries” over derivatives that had an effect on markets — as is often the case in financial markets, what matters is not the “true” underlying cause of an event, but rather what people believe so those beliefs have profound effects on prices.

        My argument is not that CDOs blew the house down, but that the whole collateralized derivative system is structurally unsound and guaranteed to lead to more severe crises that we have known in the past, whether due to losses on CDOs or some other product. The structure itself is the problem.

        “A loss is a loss.” The question in this post is whether or not the losses were related to derivatives. To the degree that you think the losses were synthetic, it appears that you agree with me that “worries over derivatives” were an important factor in the crisis.

  2. Very disturbing, to say the least. Derivatives were quite the lucrative market right up until the entire market stalled. You reap what you sow seems a little appropriate for the entire situation surrounding derivatives and how they led to the destruction of some of the largest companies in the US (until the bailouts that is).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: