Credit default swap (“CDS”) indices are now a fixture of the credit markets. Their benefits have seen them gain in popularity since their inception in 2001 and index providers continue to innovate and create products that reflect the changing economic landscape. In this article we dissect the world of CDS indices through the spectrum of the widely-referenced Markit iTraxx and Markit CDX family of CDS indices.
Credit Indices: An Overview
Synthetic credit indices have not been around for long, when compared with equities, bonds and commodities indices. They originated in 2001, when J.P. Morgan launched the JECI and Hydi indices. Morgan Stanley followed suit with the launch of Synthetic TRACERS. The two banks subsequently merged their indices under the Trac-X name in 2003. In parallel, iBoxx launched the iBoxx CDS indices. In 2004, Trac-X and iBoxx merged to form CDX in North America and iTraxx in Europe and Asia.
After administering the CDX family of indices and acting as the calculation agent for the iTraxx indices, Markit acquired both families of indices in November 2007, and now owns and manages the Markit iTraxx, Markit CDX, Markit iTraxx SovX, Markit iTraxx LevX, and Markit LCDX families of CDS indices as well as the Markit iBoxx cash bond indices.
Credit indices have expanded dramatically since their humble beginnings: Markit iTraxx and Markit CDX index trade volumes now exceed US$70 billion a day and have a net notional outstanding over US$1.2 trillion. Together they make up almost 50% of the market, when compared with the notional outstanding in single name credit derivatives. Rules, constituents, coupons and daily prices for the credit indices are available publicly from www.markit.com. Investors have also been attracted by the fact that these indices can be priced more easily than a basket of cash bond indices or single name CDS. Other advantages include the fact that they are highly liquid and are also efficient to trade due to a standardisation of terms and legal documentation. They are seen as a cost-efficient way to trade portions of the market and are supported by all major dealer banks, buy-side firms and third parties.
Markit’s CDS indices are made up of the most liquid part of the relevant single name CDS market. The Markit iTraxx Europe Investment Grade index, for example, consists of 125 European investment grade names, selected according to a number of criteria, including liquidity, ratings and underlying asset availability. With this index, an investor interested in taking a broad exposure to European corporate investment grade risk can execute this strategy with a single CDS trade rather than by purchasing 125 contracts simultaneously to make up a diverse portfolio.
Figure 1
EUROPEAN CREDIT INDICES
Structure
The indices are “rolled” every six months in March and September, a process which includes reconsidering the current constituents of the indices and issuing new indices with revised constituents. Liquidity is the main driving factor when index constituents are considered for inclusion. By using the most liquid entities traded in the single name CDS markets over the previous six months, Markit ensures that these indices are an accurate reflection of the credit markets as well as a liquid tool for all market participants.
The second major change during this “roll period” is the extension of the time to maturity by six months. The old contract—the “off-the-run” contract—remains in place until maturity, though liquidity tends to decrease as the majority of investors roll their exposure into the new contract (also called the “on-the-run” contract).
Figure 2
Trading Overview
Markit’s CDS indices can be traded, with licensed dealers providing liquidity on various platforms. Buying and selling the indices can be compared to buying and selling portfolios of loans or bonds. A buyer takes on the credit exposure to the loans or bonds, and is exposed to defaults in the same way as a buyer of a bond portfolio (buying the CDS index is equivalent to selling credit protection on the underlying index constituents). When an investor sells the index, credit exposure is passed on to another party.
The indices trade at a fixed coupon, which is paid quarterly (except for the Markit CDX Emerging Markets index which is semi-annual) by the seller of the index (buyer of protection), and upfront payments are made at initiation and close of the trade to reflect the change in price. Correspondingly, the protection seller, or buyer of the index, receives the coupon. The indices are quoted on a clean (of coupon) basis.
Figure 3
Figure 4
Spread vs. Price Indices
CDS indices are traded either in spread or in price terms. This convention mimics the bond markets, where some bonds trade on a yield basis and others on price.
Prices can be converted into spreads, and vice versa, using standardised models. Intuitively, if an index has a fixed coupon of 60 and the current coupon is 90, it is positive for the protection buyer (they are paying 60 for something that is currently worth 90). The price is inversely related to spread, so the price of the index at 90 is lower than the price at 60, and as the protection buyer is short the credit, a drop in price is positive.
Markit calculates the official levels for the Markit iTraxx and CDX suite of indices based on their regional market close times. In addition, theoretical index spreads and prices are calculated based on the contributions received for the underlying index components. These theoretical index levels are calculated using the following methodology:
The survival probability of each constituent at each coupon payment date is calculated using the Markit composite credit curve and recovery rate for each of the index constituents.
The present value (PV) of each index constituent is then calculated using the trade details of the index (as described below).
The PV of the index (Weighted Average of the PVs of the constituents) and the accrued interest on the index (Weighted Average of the accrued interest of the index constituents) are calculated.
Theoretical Price of the index is calculated as: 1+PV-Accrued.
The Index Theoretical Spread is solved as the flat curve that gives the index PV using the index recovery rate assumption.
The PV of each index constituent can be calculated using one of two methods: either a simplified model using risky duration only for each credit in the index that generates a decent approximation, or the hazard rate model for each underlying component of the index. This will generate a more accurate value, as it allows for curvature in the credit spread curve.
For small differences in fixed and current coupon the two valuation methods will have similar results. The hazard rate model will give better results for large movements in the spread. In the simple valuation methodology, the risky duration of the credit is multiplied by the difference between the current spread of the credit and the coupon of the index. This gives the PV on each component. For example, if a credit is trading at 200 bps with a risky duration of 3.75 years, and. the index coupon is 150 bps, then the PV of the constituent is 3.75*50/10,000 = 0.01875.
Total Return Credit Indices
Another way of displaying Markit’s CDS indices is to use their total return versions. In contrast to the price and spread indices, total return index levels mimic the position of an investor in the credit index market who “rolls” their position into the relevant on-the-run contract in case of a regular roll in March and September.
At a given point in time only the most recently available index CDS return is included in any one index. The return of the index therefore reflects the value of exiting the long risk position in the old Markit iTraxx contract and simultaneously entering the new contract at mid at the end of the first day of trading of the new contract. Transacting at mid means that transaction costs are not included. Therefore the roll transaction costs are 1% of the respective “old” series coupon plus 1% of the respective “new” series coupon.
By mimicking an investor position, total return indices are the only CDS indices that are linked over time. Credit total return indices are available as long and short versions of the European credit indices.
Markit iTraxx SovX: A Case Study
If an index provider had tried to launch a product based on western European sovereign CDS in 2007, dealers and investors would probably have been apathetic or incredulous. But everything changed when the global economy went through the worst recession since the 1930s. Automatic stabilisers, additional fiscal stimuli and bailouts for ailing banks left many governments with enormous budget deficits. Previously considered risk-free, many European sovereigns are now perceived by credit investors as some of the riskiest names in the world.
The sovereign CDS market has developed in tandem with the deterioration in government credit over the last three to four years. CDS on sovereign issuers such as the UK were rarely traded prior to 2007. Now the UK has the sixth-highest amount of net notional outstanding of any name—corporate, financial or sovereign (according to statistics from the DTCC). Even Italy, which tops the volumes table, was trading as tight as 5bp in the summer of 2007.
By 2009 investors were alert to the slide in sovereign credit quality in Europe. Single-name trading had picked up rapidly, and the next logical step was to create an index: the Markit iTraxx SovX Western Europe (SovX WE) was born. Comprising the 15 most liquid sovereign CDS contracts, the SovX WE allowed market participants to macro-hedge positions on European government debt and take positions on the asset class as a whole. As was witnessed after the introduction of the Markit iTraxx Europe index, liquidity tends to get concentrated in index-related trading and this drives bid/ask spreads tighter. Launched in September 2009, SovX WE is now one of the most widely traded CDS indices, typically with a bid/ask spread of about 2bp. The Markit iTraxx SovX CEEMEA index (CEEMEA), representing sovereigns in Central and Eastern Europe, as well as the Middle East and Africa, followed in January 2010.
Market participants from across the board use the Market iTraxx SovX indices to manage their risk. The bid/ask spread and liquidity of an index, compared with the underlying CDS contracts, also plays a role: entering or exiting a position cheaply and quickly is an important concern for real-money investors. The same applies to hedge funds, which use the indices for both hedging and speculation. The dealer community is obviously very active in providing liquidity.
The Markit iTraxx SovX indices have been used in a variety of trading strategies. In January 2010 the SovX WE was trading wider than its corporate equivalent, the Markit iTraxx Europe. There are caveats to the comparison. Four of the SovX WE’s 15 equally-weighted constituents are “peripheral” Eurozone countries: Greece, Ireland, Portugal and Spain. Whereas just eight of the Markit iTraxx Europe’s 125 constituents are based in these countries. The peripherals have seen a marked deterioration in their credit profiles over the last few years, and their spreads have widened sharply as result. The weighting of peripherals in the SovX WE has contributed to its underperformance.
Nonetheless, a comparison of the two indices over the last year shows that the trend has accelerated (see Figure 5). The SovX WE is now trading about 100bp wider than its corporate counterpart, a difference that can’t be explained by weighting variations alone. Some market participants will have profited from buying the SovX WE and selling the Markit iTraxx Europe, i.e. shorting sovereign credit risk and going long corporate risk. The latter index has been relatively stable since last summer, while the SovX WE has widened significantly.
Some traders have taken positions on the SovX WE and the Markit iTraxx Senior Financials. The two indices have been closely correlated—sovereigns have bailed out banks and banks are holding government debt. But this correlation has started to break down in recent months. The financials index has rallied sharply, a marked contrast to the volatility in the SovX WE. Investors are feeling more confident about many of the Eurozone banks. But the fiscal situation of the peripheral countries shows little sign of improvement, and investors are still bearish on this sector.
Figure 5
Another topic of recent times has been the rise of emerging markets and the relative decline of developed countries, when measured in terms of creditworthiness. Figure 6 shows the convergence of the Markit iTraxx SovX CEEMEA and the SovX WE over the past year. When the CEEMEA was launched in January 2010 it was trading at 221bp, more than 130bp wider than the SovX WE. A year later the CEEMEA was trading tighter than the SovX WE, a state of affairs that few would have foreseen. Again, caveats need to be stated when making this comparison. The CEEMEA isn’t equally weighted like the SovX WE: Turkey and Russia account for 30% of the index. Both of these sovereigns have performed relatively well in recent years and this has no doubt helped the CEEMEA over this period. But the main reason for the convergence in the indices was the deterioration in peripheral Eurozone credit. Investors will continue to use the two indices to reflect their views on developed versus emerging markets.
