A recent article in the Financial Times recounted that Prime Minister protest against the fact that premiums for credit default swaps (CDS) of the three largest Icelandic banks had reached astronomical levels. CDSs showing costs incurred to protect banks against default risk, the premiums induce increases in refinancing costs, which is also straining their profitability. Of course, the parties will discuss about the merits of these premiums. That said, a good understanding of the nature of discussions invites to linger about the functioning of these products.

First, CDS are part of what are called credit derivatives. These are derivatives that depend on the so-called credit events. Those events can take various forms such as bankruptcies, defaults, rating changes, and are defined with respect to the reference entities. As for the notion of derivative, it means that the payment of an asset depends on other assets or underlying variables. In the case of credit derivatives, payment depends on the occurrence of credit event. Note that, in general, the definitions of credit events are standardized by ISDA (International Swaps and Derivatives Association). However, as products often are OTC, non-standardized, all kinds of events can be mentionned in the terms of the contract. Let’s now focus more specifically on the CDS

A CDS enables a party to get a protection against a specific credit event that is the risk of default. To achieve this goal, the party in question will try to find a counterparty who, in turn, is willing to provide insurance against the specified default risk, but in return for a premium called the CDS spread. The one who "buys" insurance is called "protection buyer", while the counterpart is called "protection seller". The "protection buyer" (henceforth PB) pays a quarterly premium and the "protection seller " (henceforth PS) compensates in case of default. The value that he must pay may take different forms depending on the contract. In general, these are the two following cases. Either he pays the difference between the notional and post-default value, or he pays an amount defined in advance. With regard to the CDS contracts, several information must be mentioned:
 The underlying assets
 The definition of credit event
 The notional amount of CDS
 The start date of the CDS and the protection phase
 The maturity
 The spread of the CDS
 The series of payments (frequency)
 Payments to be made when the credit event occurs

Apart from details about the nature of contractual terms, those information define the CDS. However, even if everything seems clarified in every detail, things can go wrong. For example, what happens when the underlying asset is acquired or merges with another firm? Of course, credit risk is thereby changed. But in general, no clause specifies the impact on the CDS is in this case. It is a typical example of what economists called of incomplete contracts. To illustrate the operation of the CDS, let’s consider the following example. Two parties, the protection buyer PB and the protection seller PS, trade a CDS on Gazprom. The underlying credit is therefore Gazprom. The maturity of the CDS is 5 years. The notional amount of € 20million. The CDS premium is 120 basis points, knowing that 100 basis points equals 1%. These are the primary elements of a CDS contract, other details being generally aligned with the guidelines of the ISDA.

The premium is expressed on an annual basis, but paid on a semi-annual. The premium is expressed as a fraction of the notional. The principle of calculation is simple: just multiply the premium with the notional. To simplify somewhat, the PB will pay semi-annually (120 * 20m) / 2 = 120000EUR. This payment will continue until default’s occurence. When default occurs, the PS has to compensate for the losses incurred. In this case, various market participants are consulted to check the post-default value. Assume that the value is 430, while the notional value is 1000. In this case, the PS will pay (1000-430) / 1000 = 20m * 11.4mEUR. The PB meanwhile, may have to pay what is called Accrued fee. Indeed, when default occurs shortly after premium’s last payment, it remains an outstanding amount of premium to be paid. For example, if the premium payment was held two months ago, the PB will pay 120 000 * 2 / 6.

After this short introduction, the reader may wonder why are these products often perceived as complex. Actually, the real issue lies in computing the premium. However, to compute the premium, one must assess the likelihood of default and also potential losses from underlying credit. This is not easy and things become harder when dealing with a portfolio of credits that can be correlated. That said, even though the valuation and risk management for such products are not always easy, their potential of use is enormous and we believe that, such products will be eventually used as economic policy tools.