A Rambling and Ambivalent Defense of Credit Default Swaps
Lately, I often find myself in a position of explaining and defending CDS to my less derivatives-savvy friends and family, who are wont to heap blame for the financial crisis on the contracts themselves, the financiers who structure them, the companies that use them, and the regulators that (failed to) regulate them.
The debate usually begins in earnest when one of these friends, having read some piece in the mainstream media about CDS and the role they played in the current financial crisis, suggests banning CDS outright or worse, abrogating all existing CDS contracts. I tell them this is a bad idea and that there are legitimate uses of CDS. “Like what?” they ask.
In my usual defense, I start with the big picture. Derivatives, I tell my friends, are generally used to desirable ends by transferring (with some precision) certain specified financial risks from one party to another. By facilitating the efficient transfer of credit risk from the parties that wish to avoid them to the parties that can stomach them, CDS and other credit derivatives provide a valuable benefit to the financial system and thus to the ordinary folks on Main Street who depend on that system for personal credit (mortgages, credit cards, and other loans). I admit to them that CDS can be dangerous when improperly or recklessly used (as I believe they were at AIG), but that such danger does not warrant an outright ban on CDS precisely because of these other legitimate and beneficial uses.
This explanation usually fails to convince. I’d love to direct my friends to a more coherent defense than my own. Others have tried, including Felix Salmon who first launched his defense of CDS way-back in 2007 (and is still at it today). More recently, this teardown of CDS criticism from the wonderful (and anonymous) A Credit Trader looked promising, but I found the following statement troubling (not only because it seemed to contradict the only tenet of my simple defense):
CDS was not designed to disperse risk, per se. Instead, It was designed to do two things:
1. allow banks to get regulatory capital relief on their loan books letting them free up capital
2. allow banks to more efficiently manage credit risk (without CDS, a bank would have to sell the loans to get rid of the exposure, which is something it would be loath to do as the reference entity is likely to discover this, putting the banking relationship at risk).
I somehow think these two uses would not satisfy a CDS skeptic any moreso than my “efficient transfer of risk” explanation. For one, neither smacks of legitimacy at first blush. Further, they don’t seem sufficient to explain CDS volume, often quoted (in notional terms) in the many tens of trillions of dollars.
Why Are Regulatory Relief and Credit Risk Management Legitimate Uses of CDS?
Regarding Use #1 - Regulatory Capital Relief: A curious CDS skeptic would want to know how this “legitimate” use differs from regulatory capital arbitrage—said differently, shenanigans to circumvent the spirit (if not the letter) of the regulatory capital requirements.
To explain why this is indeed legitimate, a brief summary of my rudimentary understanding of regulatory capital requirements is in order. Banks are required to hold capital against their portfolio of risk-generating assets (net of some liabilities). The spirit of such requirements is to promote stability in the financial system by limiting the extent to which a bank can leverage it’s capital against a risky portfolio. That is, you can’t put in $1 of capital, raise $999 in deposits and other debt, and make a $1,000 in loans. The regulators would say that’s too much risk for your capital to absorb, leaving the risks to be born by the bank’s naive depositors (and gov’ts who insure them). So the regulators demand that the bank’s owners either put in more capital or shrink the portfolio.
Let’s say they shrink the portfolio. So now they’ve put in $1 of capital and have $9 in deposits and other debts, against $10 in loans. The banker then gets the bright idea to use CDS to pass-off the credit risk on the loans to a counterparty. Because the credit risk (though not interest rate risk) on the loan has been passed-off, the banker then goes to the regulator and makes a case for why he should be permitted to take on more risk (borrow money and make more loans), even though he still only has $1 in capital. The regulator agrees and the banker raises more debt to finance more lending, increasing his leverage.
This is what is meant by capital relief. Some of the banker’s capital got “freed-up” to be used to borrow more money to make more loans. It’s not nefarious and does not circumvent capital requirements because the bank has indeed hedged the credit risk. One caveat: The banker is not completely hedged. He does have some counterparty risk; but that’s a subject for another post.
Regarding Use #2 - Credit risk management with client relationship maintenance: Let me play the role of skeptic first. First, I don’t see an equivalence between selling a loan and buying protection on it. In the former, the bank has transferred all risk; in the latter, the bank has transferred credit risk, but not interest rate risk. (Granted.) Secondly, why do we need CDS to do this? If the bank doesn’t have an appetite for credit risk, why can’t they make the loan contingent on a letter of credit? Is the loan + CDS arrangement somehow better than loan + LOC? If so, why?
The loan + CDS arrangement is in many ways preferrable for both the bank and the borrower. For one, the CDS is far more fungible. If the bank later decides to accept this credit risk, he can unwind the CDS or enter into an offsetting position. For two, the CDS is a more liquid (and competitive) market. To force the borrower to shop for a LOC would have them bear some premium to compensate the guarantor for entering into such an illiquid contract. The “secrecy” benefit for the banker is admittedly less compelling.
Note also that this second use (credit risk management) is much broader in scope than the narrow example provided above. For example, counterparty risk is a form of credit risk that can be mitigated by CDS. In fact, one could argue that any legitimate use of CDS is a form of credit risk management, and vice-versa (though I defend speculation as necessary below).
CDS Intermediation: Explaining the $60 Trillion Notional
Even if you accept these uses as legitimate, you might not be satisfied that CDS as they are currently used don’t pose a danger to the financial system that warrants their increased regulation. I noted above that CDS skeptics often point to measures of volume that suggest gargantuan notional amounts outstanding. This causes them to question whether the proffered legitimate uses explain the motivation for each and every CDS.
In both of the above examples, CDS are used as hedges of a credit exposure that stems from a “direct” credit (i.e. a loan). But you can pick up the paper and read about how CDS notional volumes can dwarf the outstanding principal on a particular reference name. If CDS were designed as to be used as legitimate hedges of “direct” credit exposures, what explains the excess volume?
The answer lies in CDS intermediation (hedging indirect exposures). CDS are derivatives, not securities. They are not traded in the way that stocks or bonds are, with a piece of paper changing hands (ok, electronically changing hands). Say I’m a bank and I want to hedge a direct credit exposure such as a loan. To do so, I need to find someone willing to accept that risk. Unfortunately, these risk-takers may be hard for me to find. So I reach out to a broker-dealer who specializes in finding me a counterparty. But rather than putting me in touch with someone, the broker first takes the risk himself (creating CDS No.1) and then finds a risk-taker willing to sell the protection to him (creating CDS No. 2). End result: The bank has hedged his direct exposure, the broker has hedged his indirect exposure, and the risk-taker is the only one left exposed. We’ve now doubled the notional of CDS outstanding by introducing the broker intermediary. If the broker passes-off the credit risk to another intermediary, the notional now triples.
But… Systemic Risks
The above illustrates how CDS intermediation gives rise to a “multiplier” that can balloon the total outstanding CDS notional into an arbitrarily large amount. But this explanation gives rise to more questions. Is the total CDS notional figure really meaningless? Yes and no. The interconnectedness of a CDS intermediary to other key players in the financial system can create or compound systemic risks. Even if a CDS intermediary has hedged all of its indirect exposure, its default could (in theory) trigger a cascade other defaults among its more-exposed counterparties. That said, I think total CDS notional is a poor measure of the extent to which CDS create or exacerbate systemic risks. The significance of the systemic risks posesd by CDS depends on the specifics: who is exposed to whom and by how much? Unfortunately, untangling the web of counterparty exposures created by CDS is no easy task in the current unregulated, decentralized environment in which they are traded. This is indeed a fair criticism of the current state of the CDS market and its lack of transparency and regulation, though not of CDS themselves.
On CDS Speculation
But wait a minute, the CDS skeptic says. What about those speculators? Surely those reckless gamblers are to blame for treating CDS like a toy and bringing the credit markets to a stand-still. You’re not going to defend the speculators, are you? Yes and no.
CDS, being derivatives, are zero-sum: for every winner there is a loser. The transfer of credit risk cannot work unless someone is ultimately willing to take the risk. Enter the speculator. Funds (hedge, pension, etc.) and Special Purpose Vehicles (ultimately, the investors therein), make up the bulk of this group. AIG was another. And then there are specialist credit insurers, monolines (Ambac, MBIA) and something called a CDPC (Credit Derivative Product Company) that also speculate on credit risks with CDS.
I see no rational argument against speculation. The underlying credit risk would still exist and someone would be exposed to it whether or not the CDS market ever existed. The presence of CDS simply permits these risks to be efficiently transferred, and whoever ends up holding the risk unhedged we call them a speculator. Now some of these speculators were responsible in their risk-taking, while others were either knowingly reckless or incompetent. As with CDS intermediaries, any CDS speculator of significant systemic importance (too big or too interconnected to fail) is adding to our collective risk of systemic failure by taking an unhedged position in a CDS.
So what does this mean? If I can’t blame CDS and their originators for the financial crisis, who or what can I blame?
I hope it means that the case for banning or abrogating CDS is weak. And I’m not going to opine on the financial crisis blame game—at least not here and now.
But I do think we (ok, I) have learned a few things about CDS that we couldn’t say 18 months ago. The good news is that these lessons may have given us the insights we need to design an effective regulatory regime (or market mechanism) for mitigating the systemic risks created and/or compounded by CDS. I for one favor a CDS clearing house with sensible collateral arrangements that are blind to credit ratings. I also favor a tiered regulatory regime that imposes more stringent requirements upon entities that pose greater systemic risks.
But that’s a subject for another post.
