The Thirty Trillion Dollar Question

Do Credit Default Swaps and CDOs increase or decrease stability and systemic credit market risks?

This is not an easy question to answer, partly because we have no relevant prior experience to go by. It is such a new market and it has grown so fast that it hasn't gone through even a single full credit cycle.

Global CDS are probably around $30-35 Trillion now

I will state my conclusion right away:

Today's market arrangement intensifies existing trends. It pushes credit spreads lower than usual on the virtuous side of the credit cycle and will likely boost them higher when the trend turns vicious.

OK, I've stated my case - now I have to provide evidence. First, let's look at the current "virtuous" credit cycle.

There is no question that, until recently, credit spreads and debt defaults were at record low levels. Credit demand soared and loans were made easily and cheaply available to even the riskiest of borrowers, from sub-prime mortgages to junk loans and bonds used for LBOs. Such credit became available through the CDO's ability to pool lower quality debt, mix it with default assumptions and then slice it into large tranches rated much higher than the original debt.

Even if individual borrowers faced trouble, the rising tide of easy credit made it possible to escape default by assuming even more debt, available in the form of re-financings and equity extractions. When default assumptions are optimistic, the cascade structure of CDOs allows for a higher percentage of risky debt to be turned into immediately marketable AAA securities. Standing common sense on its head, this results in even more credit ultimately becoming available to troubled borrowers, thus avoiding default temporarily and extending the duration and reach of the virtuous credit cycle - even as it is turning into the cardinal sin of over-borrowing.

As the AAA tranches increase in size relative to the total issue, the less easily marketable pieces shrink, making their disposal easier. But they must still be sold, otherwise the whole CDO structure collapses: someone must fund and accept the risk of holding the mid-level and "toxic waste" tranches. The arranging investment banks could hold them, but this is not their function: they are best as fee-paid intermediaries, not investors. Enter the hedge funds: risk loving and performance hungry, their managers' bonuses hang on each little bit of extra return. And as they buy more of the risky tranches, once again more credit is made available to the least creditworthy borrowers and defaults are pushed down - for a while longer.

Another major type of borrower in this market is the private equity fund, raising tens of billions in loans and bonds to take over listed companies. Unlike takeovers in the 1980's that depended on direct buyers for their junk bonds arranged by Drexel Burnham, today's risky loans and bonds can be placed inside CDO's and be turned into investment grade to a great extent. During 2006, 33% of all new CDO's were backed by such risky debt (see detailed data in previous post).

There are supposed to be safety catches, of course. Rating agencies are the titular bulwarks against too much unwarranted optimism from CDO issuers. However, such agencies are paid to rate the securities by the issuer, posing huge conflicts of interest. Given the explosion of CDO deals in the past three years, rating them has become an extremely lucrative growth business and it would be going too much against human nature to be ultra conservative in granting ratings.

What about the role of CDSs? Why did such credit insurance get cheaper, creating lower spreads and making borrowing easier and less expensive? Because like any business, competition brought down price, perhaps to unreasonable levels. There is the parallel of the regular property and casualty insurance business, which often goes through pricing cycles depending on the frequency of catastrophic events. In the case of CDSs, hundreds of institutions unrelated to the insurance industry jumped into the business, eager to collect premiums. Many are hedge funds attracted by the high structural leverage inherent in CDSs and the immediate income. Synthetic and hybrid CDOs are also major sellers of CDSs, thus combining both aspects of the credit derivative and cascade elements to produce lower interest rates to borrowers.

Therefore, the current structure of credit markets lowers both interest rates and lending standards. But in no way does this mean that the ultimate, underlying borrowers become more able to properly service their debt: their credit ratings don't go up. On the contrary, by assuming even more debt they lower their creditworthiness.

Within the very CDO/CDS structure that pushed the credit pendulum from "loose" to "loosest", lies the impetus to swing it back hard and fast to "tightest".

Let's think of the process going into reverse: as more borrowers initially become delinquent, default assumptions rise and constructing new CDOs involves constructing fewer and smaller AAA tranches plus more and bigger mid- to "toxic" tranches that are difficult to sell. Credit spreads go up, CDSs become more expensive and "liquidity" dries up.

We can think of the "old" credit market structure as a lake: as water flowed in and out its level changed slowly, just inches. The current "innovative" market is a series of buckets placed one atop the other. When conditions are loose, the "toxic waste" bucket at the very top can be very small, allowing more water to spill over to the bottom AAA bucket and, being small, it is easy to hold on to it. But when more defaults occur, the size of the top, "toxic" bucket must be increased to hold the increasing "waste". Less water spills over to the AAA bucket and the "toxic" bucket becomes heavier and more difficult to hold. Since we diverted the flow to several different buckets instead of one big lake, the levels inside each bucket rise and fall much faster.

Given that no one really wants to get too wet, which system provides more stability? In the lake, many people may end up with their ankles wet as its level rises somewhat. But with the bucket scheme, those holding the top and middle ones may suddenly find themselves getting soaked head to toe. They will very quickly either ask for a wet suit (higher rates) or stop holding the buckets altogether (tighter standards).

There are some other thoughts that arise out of the nature of the CDS market as it stands today.

For example, what happens in a down market if prior CDS issuance against a single risky name greatly exceeds its total debt outstanding (keeping in mind that physical delivery of the referenced obligations is required to collect on the CDS insurance)?

It could be that, if a troubled company gets closer to default the CDS issuers (the insurers) prefer that it goes into default instead of prolonging its existence as a high-spread borrower that creates large mark-to-market losses for them. What if there is 50 times as much in CDS's outstanding than underlying debt? If the CDS's were issued evenly amongst 50 issuers, then perhaps 2% of a very bad thing (bankruptcy and default) is preferable to 100% of a bad one (very high spreads) and constant margin calls. Furthermore, won't such CDS issuers gamble that the CDS owners won't be able to obtain the bonds to put to them in order to collect on the CDS insurance? Won't that increase the market's pressure to the risky company to go into bankruptcy?

Taking the example further, say that this 50-to-1 CDS issuance is not evenly spread out, but concentrated amongst a very few issuers. In that case, those few issuers will be certain that the CDS owners won't be able collect on most of their insurance. Won't it make sense for them to write as much insurance as possible, driving this recovery percentage even lower? And doesn't that mean that the insurer will then prefer the underlying company goes bankrupt, making most CDS issued worthless? How perverse can it then be that the insured party (underlying co.) is worth more dead and bankrupt to the insurer than sick but alive? How would you like it if your medical insurance company - to whom you have been faithfully paying annual premiums - wants to pull the plug from your expensive life-support system?

Another thought; more and more CDSs are being issued, driving credit spreads lower but not, of course, raising the debt-carrying capacity of the underlying borrower. At the same time, as the CDS-to-debt ratio rises and greatly exceeds 1-to-1, the ultimate worth of those CDSs in a default event diminishes. Possible lose-lose situation: the apparent credit-worthiness of the company is increasing (lower credit spreads) allowing it to borrow more than it should - but the ultimate value of the self-same CDS's in a default event is decreased. In the end, CDS buyers paying "dirt-cheap" prices for credit insurance may be getting exactly what they paid for: dirt.

In closing, this is a complicated subject for which I am 100% sure of only this: there are many implications and possibilities that we may not see until the credit cycle turns negative.

Please provide your own thoughts and comments.