CDS: Phantom Menace

I started writing about Credit Default Swaps (CDS) just a year ago. But, for this market, that's ancient history: within just 12 months notional amounts outstanding have increased from 26 to 46 trillion dollars.

________________________________________________________________

Credit Default swaps allow hedgers and speculators to bet on the likelihood of default (or other "credit events") by borrowers like governments, corporations or pooled-asset Special Purpose Vehicles like CDOs. In addition, a very large portion of CDS contracts are now written against indices tracking debt classes such as investment grade or junk corporate bonds, MBSs, etc. That's the purpose of all those ABX, CDX, iTraxx, LCDX, etc. indices calculated by Markit.

The growth in the CDS business has been nothing short of phenomenal: within just six years amounts outstanding have increased 72-fold. The International Swaps and Derivatives Association (ISDA) in its latest semi-annual survey for the second half of 2007, put the total notional amounts outstanding at $45.5 trillion. In a separate survey, the Bank for International Settlements (BIS) reports a similar amount ($42.6 trillion) and also breaks down the instruments by type: single-name (i.e. against a single borrower like a corporation) and multi-name (i.e. against indices). The growth has been significantly faster in multi-name CDSs, suggesting increased use of CDS by speculators, instead of hedgers (see charts below).

Total CDS Outstanding (Data: ISDA)

CDS Outstanding by Type (Data: BIS)

This last finding is quite significant: even though financial index trading can and is used for broad-based portfolio hedging, it is most frequently involved in outright speculation, regardless of the instrument involved. Take the example of stock indices: it does not take much argument to ascertain that the dizzying array of broad and narrow derivatives is targeted almost entirely to speculators seeking to increase their leverage.

So why are CDSs a phantom menace, as the title proclaims? There are two major reasons:

Reason #1

At $45 trillion, the notional amount of CDS in existence is now fast approaching the total amount of credit market debt outstanding in the entire world.

Given that many debt issues are too small and unmarketable for CDS purposes (e.g. small municipal and corporate issues), it is more than likely that CDSs equal or exceed the amount of all readily marketable debt in the world. This is further aggravated by the risk concentration implied in the popularity of index trading: for example, the active CDX Investment Grade index consists of only 125 individual bonds, the CDX High Yield (junk) index of 100 bonds and the iTraxx (Europe) index of 125 bonds. While these three are not the only indices traded, the BIS survey showed there were $18.5 trillion in multi-party CDS outstanding in June 2007, an enormous amount considering the small number of bond issuers involved.

This means that far more CDSs are written relative to the amounts outstanding in individual bonds and thus credit events will infect and destroy much more speculative capital than previous increasing default cycles, when CDSs did not exist. The model is that of a viral infection or a nuclear reaction - thus their description as "financial weapons of mass destruction".

The likelihood of future corporate defaults is rising very sharply, as observed from credit spreads going up almost vertically in recent weeks: for US investment grade bonds the option adjusted spread over Treasurys has reached nearly 200 basis points and for high yield bonds 600 bp (charts below).

Merrill Lynch US Corporate Index (Charts: SIFMA)

Merrill Lynch US High Yield Index

To summarize point #1: There are too many CDSs being written by speculators against relatively few borrowers, just as the probability of default events is increasing sharply. Therefore, the possibility of a generalized financial infection through the CDS medium is substantial and rising.

Reason #2

CDSs are closely correlated to equities because, just like them, they represent business risk. CDSs have created a new way for speculators to generate highly volatile equity exposure with minimal or even zero margin requirements.

By selling a CDS on a corporation's debt, a speculator is betting that its credit will improve or stay the same. In the first case the CDS will become more valuable and result in a capital gain, while in the second the speculator will at least collect the CDS premia, almost like a series of dividends. This is identical to a speculator buying a stock expecting it to appreciate and/or pay dividends.

However, there is a crucial difference between stocks and CDSs: in purchasing a stock outright the speculator has to put up 50% margin, as required by the Fed's Reg T.

[Note: there are currently ways to lower this down to 15% through Contracts For Difference (CFDs). The Federal Reserve, much to its shame, is completely ignoring this blatant evasion. If you are familiar with bucket shop operations from the late 19th-early 20th century, CFDs are practically the same. If you are not, the classic Reminiscences of A Stock Operator is highly recommended. There are many instructive similarities today with what happened 100 years ago, despite our so-called financial innovation.]

But selling a CDS requires zero margin and even produces immediate and regular income from payments received for underwriting the credit insurance. Theoretically the sellers should maintain adequate reserves on their balance sheets to cover their credit risk exposure, but does anyone believe that hedge funds and traders actually do? Not a chance... The result is a highly volatile equity exposure, carried at zero margin in a completely unregulated over the counter market. Recipe for disaster? You bet...

Not only that: CDSs create these infinitely leveraged equity positions out of thin air. Unlike options, single stock futures or other equity derivatives that require the delivery of actual securities at settlement, CDSs do not. They are pure bubble-air and can be created regardless of the amount that is outstanding in the underlying securities.

To summarize point #2: CDSs are equity substitutes carried at zero margin, masquerading as credit instruments. They create a feedback loop mechanism to equity markets that results in reducing volatility when things look good and increasing it when they don't. In other words, they work as risk amplifiers and not as risk attenuators.

Putting the above two points together, we have the potential for a financial viral disease of pandemic proportions. The CDS market is so new that it has never been tested on the downside of the credit/business cycle. We simply have no inkling of how it will behave under real life duress, when major credit events occur with increased frequency and magnitude.

This is why I chose to describe CDS as a "phantom" menace or, as the dictionary defines it: An imaginary embodiment in threatening form of an abstract thing or quality. But in this case, the imagination is embodied in facts and the sure knowledge that the business cycle has not been abolished.

One final observation, also related to the phantom quality of the CDS market: the vast majority of people are simply unaware of its existence, let alone its importance in shaping credit and equity markets. You won't hear about it in the popular media and very little seeps through even in the financial press. This is incredible, given that it is a $45 trillion market, even if this is only the notional amount and not the value of the contracts. By comparison, the capitalization of the entire US stock market is $20 trillion.

P.S. I want to thank "cds trader" for making significant comments on the above. They appear on the comment section and so do my replies.

Something else: The term "swap" as applied to Credit Default Swaps is greatly misleading, at least in the professional meaning of the term. For example, there exist FX swaps and interest rate swaps; very large amounts of such derivatives trade OTC on a daily basis, certainly more than CDSs. Their notional amounts outstanding are even larger than CDS. But they are real swaps, i.e. the counterparties swap payments immediately and in the future. The risk involved is not at all the same as in a CDS, which is in fact an insurance policy. The only thing that is being "swapped" in CDS are regular premium payments in exchange for undertaking the risk of paying off the entire loss in case of default (equal to the notional amount minus recoveries). This is a very close equivalent to the business model of the monoline bond insurers like MBIA, AMBAC and FGIC.

The proper name for CDSs should have been DIPs = Default Insurance Policies. But that name would have certainly attracted the attention of the insurance regulators - anathema for investment banks, traders and speculators. So they called them swaps, just like the more innocent and common derivatives already residing in banks' books (off balance sheet, of course).

Oh what a tangled web...