
08 Sep 2006 - Tammer Kamel
In commenting on the series of perpetually flawed translations of his first novel, Milan Kundera noted that the whole thing would surely make him laugh if it did not concern him. This sentiment is similar to my feelings on CDOs.
A collateralized debt obligation is nothing more than a collection of bonds where all cash flows are grouped together and then redistributed. It is a mechanism for restructuring the risk/return characteristics of some set of bonds. What it actually does is shift more risk and expected return to the equity tranche while reducing the risk (and expected return) of the senior tranche.
The seductiveness of the thing is that an investor can buy an equity tranche and get triple B yields from single B bonds because they get compensated to take on default risk on behalf of senior tranches. It is so attractive because it really appears that one can get more yield than the particular credit rating warrants. For example, an investor is able to get libor + 300 basis points from a portfolio of BBB bonds or libor + 350 from the same portfolio when bundled into a CDO. This is one reason CDOs became so popular earlier this decade; interest rates were so low, investors were stretching for yield. Here was a way to extract more yield without moving out on the ratings spectrum.
This lovely feature of CDOs, by the way, has nothing to do with the choice of bonds that underlie the structure or the quality of the institution that originated the CDO. It is a pure mathematical artifact. In this sense, it is much more like leverage or the use of options: more expected return and more risk—but no free lunch. And it is the latter that many participants have not realized yet.
As the CDO market matures, there have been a number of positive developments. Most welcome is the nascent exchange traded structured credit market. The appeal of such a market is transparency of valuation. Anytime one has a visible, two way market, setting prices can no longer be the domain of a few opaque banks. In this new regime, the value of a CDO position a hedge fund holds is exactly what the market will pay for it. No need to debate models anymore.
But there is still the legacy cash CDO market. (CDOs originated by banks using real bonds as collateral and sold OTC to a set of investors.) This market is extremely illiquid. Indeed, there is often only a few parties who care what a particular issue is worth: the bank who sold it and the client who bought it. The problem here is that neither of these parties have any interest in seeing the value of the structure decrease. Since there is no market, valuations are pretty much what the bank says they are. One can not mark to market if there is no market.
It is common dictum that the cash CDO market suffered less than the synthetic and exchange traded markets last spring. Indeed, there are hedge funds out there boasting about this. The reality is, they were saved—or given a stay of execution—because no one is really marking cash CDOs properly. Since nothing technically defaulted, the interested parties can argue that their cash CDOs are still yielding what they were in February. Because there is no liquid market trading the exact structure in question, no one can “prove” them wrong.
Investors need beware of hedge fund’s holding cash CDOs right now: They may be overvalued. The fact that these funds hold utterly illiquid positions allows them (and the entire cash CDO industry) to substantially smooth out returns—valuations can’t change daily because nothing trades daily. The implication of course, is that these hedge funds are much more volatile than their returns suggest. It is disturbing, (Kundera, being unconcerned, would call it comical), that often only parties with an interest in keeping prices high are involved in "estimating" monthly mark to markets. (The originating banks and the managers.) This situation is not necessarily endemic, but it is certainly prevalent and the prudent investor treads cautiously at this time in this space.