Lynne Kiesling
Jeez, what a week … not being a financial expert, I am loathe to opine on what’s been going on in financial markets this week. I do find some things striking, particularly from a historical perspective.
The first thing that strikes me is that what’s driven this is the development and use of two new financial instruments over the past decade or so: collateralized debt obligations (CDOs) and credit default swaps (CDSs). A CDO is a slice (tranche) of a portfolio of fixed-income assets; a CDS is a contract that transfers the credit obligation for a fixed-income asset from one party to another. Investopedia has a pretty clear definition and illustration of a CDS:
The buyer of a credit swap receives credit protection, whereas the seller of the swap guarantees the credit worthiness of the product. By doing this, the risk of default is transferred from the holder of the fixed income security to the seller of the swap.
For example, the buyer of a credit swap will be entitled to the par value of the bond by the seller of the swap, should the bond default in its coupon payments.
New financial instruments like these are valuable because they allow for more, and more varied, transactions for the transfer of risk from one party to another, as well as across time. They are useful tools for hedging, and for allowing parties to be more flexible and adaptive to changes in their business environment.
Derivatives and credit instruments have existed in some form since the 12th century or so, but the size and integratedness of global markets changes the impacts when unanticipated outcomes happen. With such large and integrated global markets, the third-party rating agencies play an important role in creating healthy markets here; their evaluations are what parties use, and have used for over a century, to price risk in a transparent way. But it’s difficult to price risk, and it gets harder as the instrument gets more intricate. It’s also hard to price risk with new instruments, because pricing risk is easier when you have data on the behavior and performance of the instruments in different situations. This is why the credit rating agencies have been criticized a lot over the past year about their failure (in hindsight) to price CDO risks accurately. In terms of regulatory institutions, it doesn’t look to me like there’s much of a role for new regulation, because what’s necessary is the learning about how CDOs perform in different circumstances. But, as I said, I’m not an expert in financial ratings, so I’m interested in hearing more discussion.
Another thing I find striking is the way that the instrument transactions are consummated, particularly the CDS transactions. They are largely over-the-counter (OTC) transactions, which means that they are done bilaterally, or directly between two parties, instead of being exchange-traded. There’s a big difference in transparency, and in regulatory environment, between OTC transactions and exchange transactions, but OTC offers a lot of opportunity to create and consummate heterogeneous transactions, while exchanges focus on products that are more standardized. OTC trading clearly plays a role in enabling parties to prosper, but my question is this: does my “learning curve” argument apply here, to OTC-traded CDS instruments? I’m really not sure, and I think it’s important, because my “learning curve” argument has implications for the type of regulatory institutions we devise for these markets. Can we retain the benefits of OTC trades while simultaneously making the environment more transparent?
UPDATE: My call for help has been answered in the comments by Craig Pirrong, an excellent financial economist. In his comment he links to his post on clearing, and how a central clearing counterparty for OTC transactions could have mitigated some of these effects. It’s a superb, clear, thorough explanation, and I recommend it strongly.
Hi, Lynne–
Your intuition re the role of heterogeneity, pricing complexity, and the method of trading (OTC vs. exchange) is pretty close to the mark. Another key aspect is the relation between instrument pricing complexity, transparency, and the allocation of performance/default risk through clearing (i.e., the mutualization of this risk). I’ve done some work that points out the relation between the sharing of default risk and the need for “rocket science” pricing tools. I have also written a rather sketchy blog post on how clearing would have affected the current situation, and whether clearing is a realistic prospect for complex derivatives like credit products.
I appreciate your shining a light on this topic. It is a very underappreciated one, I can assure you. I often say that when it comes to financial markets, amateurs talk trading tactics, and professionals talk clearing (analogous to the military expression that amateurs talk tactics and professionals talk logistics.)
I think your “learning curve” argument has merit. For example in energy trading some firm specialize as, in effect, intermediaries between the OTC and futures markets – buying or selling heterogeneous contracts in the OTC world and hedging their net risk exposure to some extent on the futures market. Because a portfolio of OTC deals likely can’t be perfectly hedged by offset, the firms will end up exposed to the residual risks. While the firm likely has a model of their residual risk exposure that will help them manage their portfolio, they will only learn over time how good the model is.
In a word, yes. It isn’t derivatives per se that are bad – used wisely they can reduce risk substantially – it is using derivatives when you don’t understand the consequences. This can be something like the contract being too complex to price, but it can also be simpler things like not understanding that the realities of your market (e.g. highly non-normal price distributions) break the assumptions on which the pricing model is based, or simply not stress-testing your VaR model because you assume that “business as usual” will always happen.
If you’re referring to me, I’ve never lived in MI or TX, or been married to anyone named Tim. Mike, on the other hand … except for the Tim thing.
If you’re referring to me, I’ve never lived in MI or TX, or been married to anyone named Tim. Mike, on the other hand … except for the Tim thing.
Mike–
The paper I link to in my comment makes a similar point, namely, that “rocket science” models used for pricing and risk management services provide an information advantage to the market participants that have them. These advantages are larger, the more exotic the instrument, and the more complex the portfolio of the intermediary. I should note that the relevant private information can be of the form of (as you suggest) knowing how well the model works, particularly how well it works in guiding risk management decisions. A banker who knows his models are bad, and hence that he faces high risks because his “hedges” may blow up, has important information.
Private information impedes risk sharing. This is the nub of my argument explaining when default risk is shared, and when it isn’t. Private information is more pronounced for more exotic instruments and portfolios including larger dollops of these instruments. This implies that moral hazard and adverse selection problems (which are driven by private information) are more acute for these instruments, and that as a result there will be less sharing of default risks for them.
This general pattern is observed in practice. Even now, when there have been numerous proposals mooted to create a CCP for credit derivatives, most will commence with clearing of credit index products, rather than contracts on single names. Private information is typically less pronounced for indices.
This is also why I am skeptical that we will see a CCP in single name credit derivatives, especially on MBSs, CDOs, and the like, any time soon.
Good summaries to help understand the financial turmoil
Lynne Kiesling Last week at Freakonomics, Doug Diamond and Anil Kashyap had a great guest post to help put it in perspective, and I recommend it to all. Yesterday Steve posted a link to a critique of Paulson’s bailout plan…
Free Adult Friend Finder
Interesting topic! I really enjoyed your discussion points and plan to share your blog with my friends! I didn’t quite agree with everything but I do like to see other peoples points of view. If you want to learn how to get 100% free access to Adult…