CDS Markets – Through The Looking Glass

At the quantum level, the laws of classical physics bend in intriguing ways. In financial markets, especially at the credit default swaps (“CDS”) level, it seems the rules of finance also operate differently.

CDS contracts and credit derivatives are complex and powerful financial instruments that frequently have unforeseen consequences for market participants and the financial system. As former New York Federal Reserve President Gerald Corrigan told policy-makers and financiers on 16 May, 2007: “Anyone who thinks they understand this stuff is living in lala land.”

The derivative industry’s indefatigable support of the market centers on the fact that all the CDS contracts related to the high profile defaults in 2008 settled and the overall net settlement amounts were small.

In a previous piece – Surreal Realities of the CDS Markets –Part 1 and Part 2 ( and

) – I argued that CDS markets amplify credit losses. This has lead to some interesting debate around the effect of CDS contracts and especially their utility in financial markets.

Quantum CDS

A key part of the debate focuses on the low overall net settlement amounts. In practice, there are actually two settlements. The ‘real’ settlement where genuine hedgers and investors deliver bonds under the physical settlement rules (i.e. those who actually own bonds or loans and were hedging). Then there is the parallel universe where the dealers and large hedge funds settled via the auction. Dealers tend to have small net positions (large sold and bought protection but overall reasonably matched).

In the case of Lehman Brothers, the net settlement figure of $6 billion that is frequently quoted refers to the second process. Real CDS losses from Lehman CDS were higher, probably around $300-400 billion. Some banks and investors that had sold protection on Lehmans did not participate in the auction. They chose to take delivery of defaulted Lehman debt resulting in losses of almost the entire face value. For example, one German Landesbank reportedly took delivery of $1 billion of Lehman bonds that were worth $30 million at current market values.

There were no failures in settlement of the CDS contracts because, in part, some sellers of protection (such as banks and some insurers) were supported by governments concerned about systemic failure of the financial system. Other sellers of protection had to bear losses reducing the capital available to meet future claims. Whether the sellers are in a position to meet potential losses if default rates rise as expected remains unknown.

CDS contracts did, in all probability, amplify losses in the credit market in recent defaults. For example, when Lehman Brothers defaulted the firm had around $600 billion in debt. This would have been the maximum loss to creditors in the case of default. According to market estimates, there were CDS contracts of around $400-500 billion where Lehmans was the reference entity. The outstanding volume of CDS contracts is not known with certainty reflecting the lack of transparency about trading in the OTC market.

If used for hedging, then the CDS contracts would merely have resulted in the losses to creditors being transferred to the sellers of protection leaving the total loss unchanged. Market estimates suggest that only around $150 billion of the CDS contracts were hedges. The remaining $250-350 billion of CDS contracts were not hedging underlying debt. The losses on these CDS contracts (in excess of $200-300 billion) are additional to the $600 billion.

In addition, Lehman Brothers was included as a reference entity in other structured credit products, such as Collateralised Debt Obligations (“CDOs”) and credit indices, and additional losses may have resulted therein. Documentary asymmetries in the contracts may also increase the losses.

It is difficult to be sure about the extent about the loss amplification as there is minimal disclosure about the trading in CDS contracts as this is undertaken in the Over-the-Counter (“OTC”) market.

Loss? What Loss!

Discussions of “losses” always lead to arguments that approximate Groucho Marx: “A child of five would understand this. Send someone to fetch a child of five.”

Proponents of CDS contracts argue that losses on additional “speculative” positions on Lehman as a reference entity are not a loss per se; that is, there are no “real” losses. Instead, the argument goes that they represent a transfer of payments from one counterparty to another – from the seller of protection to the buyer of protection. Derivative advocates then go on to argue that these “speculators” also “benefit” the market through increased liquidity and greater price discovery.

The true issue is subtler – the CDS contracts amplified the losses as a result of the bankruptcy of Lehmans by (up to) approximately 50%. It increases the embedded leverage in the financial system to a specific event namely the default of the reference entity. It also may absorb available liquidity and capital creating systemic issues.

The argument regarding transfer between buyers and sellers of protection is correct but merely confirms a truism; i.e. that no wealth is ever made in financial markets it is just transferred between market participants.

Interestingly, if you follow that logic with Lehman’s debt then nobody actually lost money either because somebody somewhere must have made it on the other side. Indeed, there have been no losses at all during the entire Global Financial Crisis as someone somewhere has made the offsetting gain. The pain, suffering, bewilderment and powerlessness of millions of individuals who have lost their livelihoods, savings and futures are, under this argument of course, not real either.

The attitude of self serving Masters of the Universe and their fawning acolytes in this regard is reminiscent of a French Queen who once observed of her subjects: “Let them eat cake.”

“ID” or Innovation Dysfunction

Financial innovation can offer economic benefits. A number of major benefits of CDS contracts are often cited by academic acolytes and fans, generally those promoting the product.

The first is that CDS contracts help complete markets, enhancing investment and borrowing opportunities, reducing transaction costs and allowing risk transfer. CDS contracts, where used for hedging, offers these advantages. Where not used for hedging it is not clear how this assists in capital formation and enhancing efficiency of markets.

CDS contracts also, it is claimed, improve market liquidity. It is generally assumed that speculative interest assists in enhancing liquidity and lowers trading costs. Where the liquidity comes from leveraged investors, the additional systemic risk from the activity of these entities has to be balanced against potential benefits. The current financial crisis highlights these tradeoffs.

CDS contracts also, it is claimed, improve the efficiency of credit pricing. It is not clear whether this is actually the case in practice.

Pricing of CDS contracts frequently does not accord with reasonable expected risk of default. The CDS prices, in practice, incorporate substantial liquidity premia, compensation for volatility of credit spreads and other factors.

CDS pricing also frequently does not align with pricing of other traded credit instruments such as bonds or loans. For example, the existence of the “negative basis trade” is predicated on pricing inefficiency.

In a negative basis transaction commonly undertaken by investors including insurance companies, the investor purchases a bond issued by the reference entity and hedges the credit risk by buying protection on the issuer using a CDS contract. The transaction is designed to lock in a positive margin between the earnings on the bond and CDS fees. Negative basis trades exploit market inefficiencies in the pricing of credit risk between bond and CDS markets.

In early 2009, the pricing of corporate bonds and CDS on the issuer diverged significantly. For example, the CDS fees for National Grid, a UK utility, were around 2.00% pa (200 basis points) compared to National Grid’s credit spread to government of around 3.30% (330 basis points). Similarly, Tesco, the UK retailer was exhibited CDS fees of around 1.40% (140 basis points) against a credit spread to government of around 2.50% (250 basis points).

In effect, market pricing of credit risk as between the CDS market and the bond and loan market was significantly different.

Another area of pricing discrepancy is the relative pricing of different firms. For example, in early 2009, bonds issued by borrowers rated “A” were trading at a higher credit spread than bonds of borrowers rated lower (say “B”) in the bond market. At the same times, CDS fees for borrowers rated “A” were trading at a lower level than CDS fees of borrowers rated lower (say “B”) in the credit derivatives market.

There are also notable discrepancies in the pricing of corporate credit risk relative to their sovereigns. In early 2009, Cadbury, the UK confectionery firm, was trading for 10 years substantially below the CDS fee of the UK government but Cadbury bonds were trading at a spread of around 2.00% (200 basis points) above UK government bonds. As people on one side of the Atlantic Ocean might remark: “Go figure!”

CDS contracts also are supposed to enhance information efficiency, improving availability of market prices for credit risk allowing more informed decisions by market participants. As CDS contracts are traded in the private OTC derivative markets, there is limited dissemination of market prices. This limits price discovery and therefore any informational benefits.

In reality, pricing and trading information is only available readily to large active dealers in CDS contracts. This informational asymmetry may advantage these dealers. Knowledge about trading flows in CDS contracts may allow these dealers to earn economic profits.

Benefits of CDS contracts must be balanced against any additional risks to the financial system from trading in these instruments. CDS contracts may create additional risks within the financial system. While CDS contracts did not cause the current financial crisis (excessive reliance of debt did), they may have exacerbated the problems and complicated the process of dealing with the issues.

As John Kay noted (in “The Long and the Short of It” (2008) Erasmus Press at 222): “The academic cheerleaders were, in the main, innocent dupes…they provided philosophical justification for greed and rightist ideology, through a Panglossian message of [efficient] markets…”

Risk Reduction or Risk Creation?

The CDS market originally was predominantly a market for transferring and hedging credit risk. The contract itself has many attractive economic features and can serve useful purposes in hedging and transferring risk. Even this hedging application is dogged by some of the identified documentary issues that may reduce the effectiveness of CDS contracts as a hedge. Such problems may well be fundamental to the nature of the instrument and incapable of remedy, at least easily.

In recent years, the ability to trade credit, create different types of credit risk to trade, the ability to short credit and also take highly leveraged credit bets has become increasingly important. To some extent the CDS market has detached from the underlying “real” credit market. If defaults rise then the high leverage, inherent complexity and potential loss of liquidity of CDS contracts and structures based on them may cause problems.

The International Swaps and Derivatives Association (“ISDA”), the derivatives industry group, have recently implemented initiatives to “hard wire” the auction based protocols into the standard CDS documentation. They have also initiated changes in market practices, such as fixed coupons for CDS contracts, designed to facilitate trading in these instruments. These actions increasingly focus on CDS contracts as an instrument for trading on default risk and credit spreads rather than one whose primary objective is the hedging of credit risk. The latter would emphasis less standardisation and a greater focus on matching the nature of underlying bond or loan being hedged.

The excesses of the CDS market are evident in the recent interest in contracts protecting against the default of a sovereign (known as sovereign CDS). For example, the CDS market for sovereign debt is increasingly pricing in increased funding costs for the US. The fee for hedging against losses on $10 million of Treasuries currently is about 1.00% pa for 10 years (equivalent to $100,000 annually). This is an increase from 0.01% pa ($1,000) in 2007.

The spectre of banks, some of whom have needed capital injections and liquidity support from governments to ensure their own survival, offering to insure other market participants against the risk of default of sovereign government (sometimes their own) is surreal.

The unpalatable reality that very few, self interested industry participants are prepared to admit is that much of what passed for financial innovation was specifically designed to conceal risk, obfuscate investors and reduce transparency. The process was entirely deliberate. Efficiency and transparency are not consistent with the high profit margins that are much sought after on Wall Street. Financial products need to be opaque and priced inefficiently to produce excessive profits or economic rents.

In October 2008, Alan Greenspan, the former Chairman of the Fed, acknowledged he was “partially” wrong to oppose regulation of CDS. “Credit default swaps, I think, have serious problems associated with them,” he admitted to a Congressional hearing. This from the man who on 30 July 1998, stated that: “Regulation of derivatives transactions that are privately negotiated by professionals is unnecessary.”

On 6 March 2009 Bloomberg reported that Myron Scholes, the Nobel prize winning co-creator of the eponymous Black-Scholes-Merton option pricing model, observed that the derivative markets have stopped functioning and are creating problems in resolving the global financial crisis. Scholes was quoted as saying that: ” [The] solution is really to blow up or burn the OTC market, the CDSs and swaps and structured products, and … start over…” ISDA, the beleaguered derivatives industry group, predictably countered limply that: “… the notion that you would, as he said, blow up, the business in that way is just misguided.

As the global economy slows and the risk of corporate default increases, it remains to be seen whether CDS contracts act as instruments of risk reduction or amplify risk. Recent debates about and regulatory proposals to regulate CDS markets show a disappointing and limited awareness of these issues, most worryingly amongst people who claim to know!

© 2009 Satyajit Das All Rights reserved.

Satyajit Das is a risk consultant and author of Traders, Guns & Money: Knowns and Unknowns in the Dazzling World of Derivatives (2006, FT-Prentice Hall).

9 Responses to "CDS Markets – Through The Looking Glass"

  1. NFrazier   March 23, 2009 at 7:56 pm

    In the Ising model, if a lattice of interacting magnetic moments is above a certain critical temperature, heat causes the directions of the magnetic moments to be uncorrelated. If the temperature drops below this critical point, then the directions of the magnetic moments become spontaneously correlated. In thermal mechanics, this mechanism is called a phase transition. In statistical quantum field theory, it’s called “spontaneous symmetry breaking.”Similarly, the idea of reducing risk via CDS’s is valid as long as the “temperature of the economy” is sufficiently high to keep the default of the debts within it uncorrelated (when there isn’t much risk anyway…) and/or there are no latent interacting channels of transmission (economic, financial, etc) between the debts in the CDS. If “the temperature of the economy” drops too low and these latent channels are activated, then the individual debts of the CDS become spontaneously in default. In finance, this is an example of a leptokurtic event. The economic analog of this would be (the reversal of) a multiplier effect.Warren Buffet was right to call these instruments “weapons of mass financial destruction.” That so many ignored the fact that these instruments were not immune to spontaneous auto-correlation created a false sense of security and greater investment in them, and hence actually increased risk. Imagine further what would happen if a substantial portion of a major firm’s income came from selling deep out of the money put-options on CDS’s. That would leverage the losses under a spontaneous auto-correlation to catastrophic levels; possibly enough to sink the global economy.Theoretically, the “phase transition” could work in the other direction as well (for CDS’s that don’t already have irreversible foreclosures within them anyway). So it is easy to see why some are in a panic to boost the value of these instruments, jawbone some confidence back into them, and get the economy going again.Such a depressive-manic switch would split off the fundamental reality that a large part of the global economy is structured around unsustainable debt-financed consumption in the US however. This raises the fundamental question of whether the banks should just be rescued in the short run or in the long run as well. The answer to this question should rely principally on how many profitable entrepreneurial opportunities are out there in the real economy – and not on how to postpone an inevitable restructuring for years until remaining lines of credit are gone and policy tools are broken.

    • Guest   March 27, 2009 at 12:38 pm

      The entrepreneurial opportunities do not exist to the degree to pull us out. This is why the Fed intervened in the early 90s to begin with creating the 15 year artificial economy. In the meantime, the few opportunities that did exist were not invested in but outsourced. Government should have used capital acquired in the boom to invest in production and make the US more entrepreneurial and competitive (reduce regulations, taxes, etc.).

    • kiers   March 28, 2009 at 1:05 pm

      this is what happens when I-Bankers take years of science courses….they cease to communicate in english. brandishing their brainpower.In plain english:Price of CDS= (prob of default) x (depth of default).However, the mere existence of CDS INCREASES the (prob of default)==> illogical to price a circular formula

      • Guest   March 31, 2009 at 8:24 am

        Using derivatives (such as CDS and CDO) and the scientists behind their creation as escape goats is very narrow minded. Had you read with depth the excellent article of Mr Das, you would see that he remarks how the original idea behind CDS was hedging against credit default, and that a distortion from hedging to speculating scenarios was at the root of the problems (e.g. Lehman Brothers and AIG).Of course, under the current circumstances a Frankenstein’s approach is tempting, but I would rather remind you the difference between science and engineering: the nuclear power was a great discovery to humankind, whereas Hiroshima and Nagasaki was not Einstein’s fault, better blame the army.Same debate between stem cells research and ethics, and the very same happens with finance: the inexistent standards of lending (e.g. subprime), the reckless rush for higher yields, the misunderstanding of the basics of derivatives pricing, the misuse of VaR and copulas… those are the ones to blame.

  2. interested reader   March 23, 2009 at 10:07 pm

    What would you suggest is the likely endgame?Thanks for your insightful analysis.

    • Franco   March 27, 2009 at 7:56 am

      The endgame will be determined by simple facts; its surprising how people keep calling a bottom.When jobs are being lost, spending cannot recover, new homes cannot be bought, and demand for manufacturing is lower.I do not even agree that markets discount a recovery very much in advance…its just nominally so;Go back to 2003, see the ISM data, jobless claims and new home sales etc.the market will start rising when:The ISM starts rising for abt 3 months in a row and approaches 50.Non-farm payrolls get nominally +Personal income and spending approaches +0,3Philly fed is above 15Factory orders are positiveNew home sales > 950 andindustrial production is +2% or more.By any standards we are EXTREMELY FAR from the above.

  3. Paul94611   March 23, 2009 at 11:42 pm

    Whatever engineers can construct nature can destroy.Taken in a slightly different way by Timothy Geithner when discussing his much awaited program as reported by the Financial Times: for Mr Geithner, he was choosing his words very carefully on Monday. “We need to balance the basic objective that we do not reward failure against the hugely important imperative that we get the financial system doing what it needs to do,” he said.I suspect that as Satyajit Das discusses the primary areas of concern in dealing with the “financial crisis” will not be found in bank “assets”, but in their “liabilities”. Especially when so many institutions like AIG created 10X or more in liabilities from one asset through the CDS marketplace.To this old retired knuckle dragger it appears as if the proponents of the CDS still subscribe to the notion of a steady state universe. Or, as one of Murphy’s Laws of combat clearly states; “Tracers work in both directions”.

  4. DM   March 27, 2009 at 9:26 am

    I have my doubts that CDS contribute to efficient pricing of risk although how exactly they price this stuff is a mystery to me and something I hoped you would have better addressed.I think you miss the more fundamental point that it is very hard to price credit accurately. Low probability events and enormous reward asymetry is not easy to get right. So while CDS can be all over the logical map – as you correctly point out – the guys pricing the bonds and loans in the last few years didn’t do such a great job either.

  5. Guest   March 27, 2009 at 10:19 am

    Satyajit Das is an excellent writer. Would like to see him write weekly or even more often on what is happening in the marketplace. A daily financial blog by him similar to Mish Shedlock’s would be most welcome!