One of the downsides of the Global Financial Crisis (GFC) has been the tarnished reputation of certain financial instruments that were developed with good intentions, namely derivatives. While there is no disputing that the misuse of these products exacerbated the turmoil of the GFC, the original rationale for their existence was to enable effective risk management in financial markets. Credit Default Swaps (CDS) were one such instrument. As this paper explains, when used appropriately, CDS play a vital role in credit markets as they allow portfolio managers to address market and credit risk in an effective manner. Additionally, recent regulatory changes mean that many of the issues pertaining to the misuse of CDS instruments immediately prior to the GFC are unlikely to be repeated.

Financial derivatives have often been portrayed as recent innovations, yet there is evidence to suggest that put options existed in the early human civilisation of Mesopotamia. It is hypothesised that farmers, by decree of the King, had their debts forgiven if there was insufficient rain. The modern finance industry has been responsible for industrialising many of these instruments. In the lead-up to the GFC, financial engineering led to the abuse of many instruments, all of which were developed with the intention of allowing parties to mitigate risk in a similar manner to the Mesopotamian farmers.

The principles of risk management

The role of an investor is to deploy their capital to those in need of capital. In return for their capital, an investor demands a return. At a simple level, an investor can perform their role by either loaning capital to an entity via a direct loan, purchasing a bond, or obtaining equity in the entity by purchasing shares. A share entitles the investor to a share of profits, via a dividend, and last claim over the assets of the entity. The expected value of these components is captured by the share price of the firm. If an investor wants to protect the downside to their equity investment, they can either sell some shares, assuming there is a market, or acquire a put option. A put option means that a counterparty has agreed to acquire a share at a predetermined price. Therefore, the investor effectively removes any downside risk if profits fall, or the company fails.

Alternatively, an investor can be a debt financier. In return for providing their capital for a determined period, either through a direct loan or purchasing a bond, the investor will receive a regular coupon and have their original capital returned at the maturity of the loan. In this instance, if the investor feels the probability of them not receiving all their coupons and/or not receiving the full value of their capital back is too high, they can sell their investment. However, debt markets aren’t generally as liquid as equity markets. An alternative is to find a derivative that can mitigate the risk. In this case a CDS is an appropriate tool for this role.

CDS can also play an important role in the construction of a fixed income portfolio. For example, credit managers wanting to manage the credit duration (the effect of a change in credit spread on an asset’s value) of their portfolio can utilise CDS to adjust this duration by either going long or short credit.

A logical question is, if an investor wanted to go short credit, why would not they simply sell their debt instruments? The answer involves the concept of a portfolio overlay. An overlay is the process of using a derivative, such as CDS, to obtain, offset or substitute specific physical portfolio exposures. There are multiple scenarios where an overlay is appropriate, including:

• Investors may see a short-term, low probability risk they want to hedge against without liquidating their position in an otherwise well performing asset. An example of this scenario was the recent near calamity over the US debt ceiling.
• Credit spreads are expected to move wider over the medium as the business cycle matures. If a manager is holding well performing assets, they may find it difficult to replace them. Therefore, it would be effective to hedge those positions.
• A given market has become illiquid, that is the buy/sell spreads have become excessive, so it would be more cost effective to add the portfolio overlay.
• In most cases a CDS will have a higher beta, in that its price will move by a greater degree than the reference entity. Thereby, allowing CDS to provide a cost-effective hedge.
  
  

The birth and near death of CDS

J.P Morgan introduced CDS in 1994 for the purpose of allowing banks to protect themselves against defaults losses within their loan or corporate bond portfolios. A straightforward solution for reducing default risk would have been for the capital providers to sell the relevant security. However, this step might see the capital provider impair their relationship with the borrower, which could lead to reduced opportunities to lend to them going forward.

In basic terms a CDS is an insurance policy, whereby the buyer of CDS is looking to protect themselves against losses arising from a default event. Section 3 details how CDS allow banks, or any other lender, to separate the default risk on loans from the loans themselves. The risk of default faced by the lender is influenced by time, company specific factors and general economic conditions. The price of protection using CDS is constantly adjusting as the probability of a default is influenced by the aforementioned variables.

Unlike traditional equity securities, the original CDS were traded over the counter (OTC) via bi-lateral agreements. Given the agreements related to a single entity or a particular set of loans, they were known as single-name CDS. The original process saw one party look to offload risk by buying protection via CDS, while the party who was willing to accept the risk would be the seller of protection via CDS. The acceptable market jargon to explain the transfer of risk is that the seller is going ‘long credit’ while the buyer is going ‘short credit’.

From humble beginnings, the global CDS market underwent rapid growth in the mid-2000s – the outstanding notional value of all contracts grew from $6.4 trillion in 2004 to $58.2 trillion by 2007. A significant proportion of this growth came about due to speculators, including hedge funds and relatively vanilla mutual funds, utilising CDS to enhance their returns rather than to simply eliminate default risk. This trend saw speculators selling CDS, thereby going long credit as they assumed benign credit conditions and wished to collect the premium payments. This strategy ultimately proved extremely costly at the height of the GFC.

A direct effect of the epic growth in the issuance of CDS was the pricing of default risk. With speculators looking to go long credit, in essence this created an oversupply of CDS, which suppressed credit spreads well below the corresponding spread on the physical asset. This outcome in turn drove additional interest in the CDS market, perpetuating the issuance of CDS right up the onsite of the GFC.

The poster child for the excessive use of CDS was synthetic collateralised debt obligations (CDOs); a product that utilises CDS, rather than a physical pool of assets, to create a CDO. The main upshot from the introduction of synthetic CDOs was a dramatic increase in unseen leverage, as speculators placed multiple positions on the same pool of assets.

Upon the single name CDS becoming widely adopted, regional CDS indices (see Section 3.3) were introduced to cover both investment grade (IG) and high yield (HY) markets. These indices quickly established themselves as the most liquid instrument in the credit markets. Table 1 summarises the main credit indices available in today’s market, with the bracketed number indicating the number of single-names CDS included in the index. The table also provides the tenor of each index, with 5 years being the most common tenor. Section 3.3 provides more detail on the relevance of tenor for the CDS contract.

In the aftermath of the GFC volumes in the CDS market collapsed, though in recent years they have recovered somewhat with the total notional outstanding at end of 2022 being $16 trillion. The collapse in CDS market volumes reflected the significant amount of trading immediately pre GFC that was undertaken by market participants with little or no exposure to the assets underpinning each CDS contract – that is they were placing “naked bets.” This activity was summed up by Warren Buffett, who labelled CDS as a “financial weapon of mass destruction.” The basis of Buffett’s criticism was that, as discussed earlier, CDS became a tool for massive speculation. Originally, speculators ran rampant selling CDS before switching, and bought CDS to achieve a similar outcome as naked-short selling ^[1], placing downward pressure on a host of securities.

A consequence of the OTC nature of the CDS market, including no independent trade clearing process, was the lack of transparency and regulations. The direct result of this was that the parties were simply unaware of the counter-party risk that existed in the market, a matter amplified by the popularity of synthetic CDO. Counter-party risk refers to the risk that the other party in the bi-lateral agreement cannot meet their commitment. Of particular concern was that the sellers of the CDS would have insufficient funds to pay out in the event of a credit event. The specifics of this are explained in Section 3.1.

The significance of the previous point was made clear when global insurance company AIG required a desperate bail-out as defaults escalated across the globe. AIG’s bail-out was required because they had lost track, or were unaware, of their massive exposure to failing mortgage-backed securities (MBS) and CDOs, plus a variety of other derivatives where they had sold protection. Another issue for AIG was that they had completely mispriced the risk instruments they sold as they failed to anticipate any deterioration in default rates.

In very simplistic terms, the AIG situation was akin to an insurer not being able to meet all claims resulting from a major earthquake hitting San Francisco; a situation which would place the capital reserves of all insurers under intense pressure, but particularly those that had chased additional premiums by sacrificing risk. If this situation did occur, would people seek the end of the home insurance market? Likely not; the more sensible approach is to adjust market practices to prevent similar events in the future. As discussed in Section 4, many of the shortcomings of the CDS market have been addressed and the market is now operating with a framework the removes counterparty risk and offers standardised contracts.

Mechanics

The theory

Putting aside speculative behaviour, predating the origination of a CDS contract an investor would have gained physical exposure to a reference entity (a company or loan). To mitigate the default risk associated with the physical exposure, an investor will purchase a single-name CDS linked to the reference entity. Additionally, with the advent of CDS indices an investor can increase, or reduce, their exposure to a group of single name CDS (see Section 3.3). The underlying mechanics of these CDS indices are consistent with single-name contract.

Figure 1 provides an overview of the parties involved in a bi-lateral CDS contract. As stated earlier, CDS are a form of financial derivative, whereby a derivative is a synthetic version of a physical transaction. Therefore, the seller of a CDS, the party going long credit, becomes a synthetic lender to the reference entity. The seller is thereby assuming the reference entity will incur a credit event which would force the payout to the CDS buyer. Critically, the contract does not run in perpetuity, rather the agreement will remain in place for a finite period. This period is known as the tenor of the contract, and per Section 3.2 plays a vital role in the pricing of CDS.

There are three possible payment components to CDS contracts. The first two occur irrespective of a credit event (default), with the third dependent on a reference event.

Upfront payment

Per Figure 1 the first payment occurs at initiation and involves an upfront payment by one of the parties. Post the standardisation of CDS contracts in 2009, this payment represents the difference between the newly introduced standardised quarterly premium (for example, in the US this is 1% for IG credit and 5% for HY credit) and the price of purchasing protection on the reference entity for the tenor of the contract. The latter component is effectively the credit spread for the reference entity. The notional value of the protection being sought also affects this payment.

Section 3.2 provides more detail on this initial payment. In summary, the price of protection considers the perceived risk of the reference entity defaulting, which in turn is affected by the tenor of the protection. The notional value relates to the value of the protection in the event of a credit event. A credit event is an event that allows the buyer to “make a claim” on their contract.

Premium leg

The next payment is the premium leg. This payment takes the form of a quarterly coupon by the buyer and is akin to an insurance policy payment. The buyer makes these payments across the tenor of the contract. With the standardisation of CDS contracts, the premium leg involves a fixed coupon determined by the credit quality of the reference entity. Section 3.2 discusses this issue further.

In a benign environment, this premium coupon looks very attractive to CDS sellers as there is an unlikely (perceived) risk of having to pay out on any contracts. Therefore, the seller is likely to recycle the incoming coupons into other investments. Indeed, this is the trap that AIG fell into, with the outcome compounded by AIG not pricing risk appropriately. Therefore, when it was time to make good on their swaps, AIG had greatly underestimated the funds required to meet their obligations.

Contingent leg

While the bankruptcy of, and payment default by, the reference entity are the main credit events that a CDS buyer seeks protection against, per Table 2, others do exist. If one of these events occurs then based on the terms of the contingent leg, the CDS seller will need to pay-out to the buyer.

Two options exist for settling a CDS if a credit event occurs. Figure 2 presents the simpler option, which is a one-way payment from the CDS seller to the protection buyer. This payment is contingent on the recovery rate of the underlying capital. Within the pricing formula for standardised contracts there is an assumed recovery rate that varies based on the credit quality of the reference entity, An IG CDS contract assumes a 40% recovery, while HY rates are set dynamically.  However, this assumed rate is not a guarantee. Once a credit event occurs a ‘calculation agent’ is appointed to establish a specific recovery rate. This process involves the calculation agent sourcing quotes on the defaulting bond, with this outcome forming the basis of the settlement. Often the calculation agent is the seller of protection.

The other settlement method involves each party delivering either a payment or a security. If the protection buyer is “naked” they will need to acquire the defaulting security so they can deliver the physical security to the protection seller. Given a credit event has occurred, this value will be well below the securities original par value. In return, the protection seller will deliver the notional value of the CDS. In the instance the buyer holds the asset they would transfer the bond and receive the notional value of the CDS.

Pricing

While CDS pricing is highly mathematical, it is relatively straightforward to isolate the significance each component contributes to the “traded rate” or, in simpler terms, the spread. The main driver is the perceived default risk of the reference entity, or entities in the case of an index. This risk will fluctuate with general economic conditions and issues specific to the reference entity. The rate is also influenced by the tenor of the contract, with longer tenors generally attracting higher rates as the probability of default increases with time. This traded rate affects the initial upfront payment and the ongoing mark-to-market value of the contract.

The CDS market maker offers different rates to the market to buy or sell protection for a given reference entity. For the seller, who is going long credit, the higher the perceived risk, the higher the rate at which they will wish to offer protection. Alternatively, for the buyer, they will want to acquire protection at the lowest possible spread. Therefore, a buyer may not want to wait until a default is imminent to purchase protection as the spread will be prohibitive.

An important aspect of the OTC market is the limited transparency of pricing compared to an equity market. In most equity markets, traders can observe the prices that each share parcel of a particular company is sold for and the current trading depth of the market. However, in a manner consistent with the broader fixed income market, and given the OTC nature of the CDS market, a buyer can only see the most recently posted quotes from each of the CDS sellers for a given contract. A buyer may receive either more or less favourable terms depending on how the market has changed since the publishing of the last quotes. The buyer then decides to accept the rate offered by the seller or seek a better rate from another seller.

Once a contract is struck, its value is determined as the difference between the rate at which it was struck versus the latest pricing for equivalent contracts^[2].  Like other financial instruments, the value of a CDS contract will not remain static. CDS buyers (a short credit position) will have the value of their swap decrease if the rates for new contract have fallen, indicating the market’s assessment of default risk has fallen. This result occurs because protection has become cheaper to buy and therefore any contract with a higher rate is less valuable. Alternatively, as credit spreads widen, the buyer will benefit as it would cost more to arrange the same protection.

The above pricing dynamic explains how a credit manager can utilise CDS contracts to hedge. That is, while an adverse credit event, or market conditions, negatively affects the value of their physical assets, the events will positively affect their synthetic portfolio.

Indices

As mentioned in Section 2, as the CDS market matured, various bodies introduced CDS indices. Like an equity index, CDS indices allowed investors to gain exposure to a basket of underlying CDS rather than having to pick and choose single name instruments. Per Table 1, the indices cover a range of geographies, tenors, and credit ratings, thereby providing the flexibility to investors to take a view on certain sections of the market. Additionally, the presence of the indices provided greater liquidity and tradability for those in the credit market, with the bonus of lowering transaction costs and increasing transparency. In terms of liquidity, only the most liquid names are included in CDS indices, thus ensuring that the indices remain liquid and potentially more stable.

There is one significant difference between an equity index and a CDS index, which is that CDS indices roll every 6 months. The rolling process involves creating a new index as the constituents of the index are reviewed, and the new index/contract formed. Despite the new index, the existing contracts do continue to trade, albeit with slightly reduced liquidity and are deemed to be “off-the-run”, compared to the new “on-the-run” issue.

Clearing up the old ways

In the aftermath of the GFC, it became clear that the opaque nature of the CDS market needed to improve. Underpinning this need was the importance of the CDS market in assisting to appropriately price risk across all asset classes. One of the major changes was the introduction of more formalised clearing. The intention of this change was to remove the counterparty risk, as seen with AIG, and therefore restore confidence in the market. The other change was the introduction of standardised contracts.

The most significant change in the clearing process was the introduction of International Exchange Clear Credit (ICECC). Unlike an equity clearing house, which matches buy and sell orders, clearing a CDS requires each party to post collateral in the form of cash or another highly liquid asset to ensure the contracted payments, monthly coupons, or default payments, can be made. The role of the ICECC was to step in as a buyer to match every CDS seller and alternatively be a seller for every buyer. ICECC was set up with multiple layers of protection, including parties posting collateral, to ensure all contracts would be honoured. This process took away the counterparty risk, leaving default risk of the reference asset as the main determinant of pricing.

Conclusion

While financial derivatives are not new, there is little doubt that at times they have been abused and caused excess volatility. One of the many outcomes from the GFC was that investors came to see CDS as recklessly speculative financial instrument. However, the underlying intention of CDS remains consistent with the prudent management of a credit portfolio. That is; a manager wishing to reduce their credit exposure in the face of an anticipated increase in market volatility can utilise CDS for this role. With recent changes to clearing and greater transparency, the CDS can again be legitimately considered a vital part of modern financial markets.