Liquidity Signals In The CDS Markets

Enhancing CDS monitoring with OTC specific quoting patterns

MICHAL KOBLAS, HEAD OF QUANTITATIVE RESEARCH, CREDIT MARKET ANALYSIS
Originally published in the May 2010 issue

On the day when KfW wired €300 million to the defaulted Lehman Bros, it became clear that a new regime for risk control and counterparty risk assessment was imminent. No longer could the middle office operate in an end-of day or end-of-week environment whilst the front office operated in real-time. There was a requirement to extend dynamic environments from the front office into departments responsible for managing counterparty credit risk.

This article illustrates how an institution can significantly enhance their ability to actively manage their counterparty credit exposures by using CDS market activity information. This article will also introduce CMAs market activity indicators, which provide CDS market information that is not contained in CDS price levels but can have a significant and valuable impact on counterparty credit assessment.

Evolution of credit risk assessment
For a long time, counterparty credit assessment was largely outsourced to the big three credit rating agencies and was subject to a rather static approach in its implementation. Recent credit market turmoil exposed some potential shortcomings of this practice. Credit ratings alone simply did not satisfy the increasing need for enhanced transparency and close to real-time risk monitoring. Their long credit evaluation cycles could not provide a timely and responsive flow of relevant information to market participants who had found themselves holding large exposures on rapidly deteriorating credits. So institutions that needed an up-to-the-minute picture of credit risk began to look beyond the rating agencies for other methods to assess their credit and counterparty exposures.
In the pursuit for a more timely and accurate assessment of their credit and counterparty risks, institutions turned their attention to the credit derivatives markets. Here the strength of individual credit is evaluated on a close to real-time basis and opinions are backed by the capital exposure of those who buy or sell exposure in the reference credit. While the limited outstanding debt notionals in the corporate bond market often mitigate its pricing efficiency, the CDS market, with its ability to write protection on gross notionals multiple of the reference entity’s outstanding debt, became the natural venue for firms who needed a responsive and reliable source of credit assessment information.

Understanding CDS market
Due to the OTC mechanics of the CDS market, the ability to access reliable pricing data and trading information is a challenge for market participants. In contrast to exchange-traded markets, the mechanics of CDS data collection plays a significant role in the value of the information that may be extracted from CDS market data.

While genuine transactional data (so called Level 1 pricing) is available for only a very limited number of single name credits, trading of a majority of single name credits is transacted without any records being made easily accessible. Although a few CDS data vendors provide aggregations of book and record figures, these figures lack a direct connection to CDS front office activity. Assessment of the reference credit then relies on using CDS price figures alone, without transparency on the underlying market activity. So Level 3 data may not say much aboutthe market’s perception of the underlying entity’s probability of default, or provide comprehensive indications of its potential evolution.

Level 2 CDS pricing data, available from CMA, originated in a close to real-time flow of executable price quotes that buy-side market participants receive from their market makers. The data is one of the most commonly used market-based pricing data sources, providing a much more accurate insight into the CDS market’s view on an individual entity’s credit standing.

In order to fully comprehend the market dynamics surrounding specific reference entities, it is important to monitor the price of protection in conjunction with information on the size and scale of market activity around a particular credit. By monitoring sudden drops or rises in market interest in a given credit, one can gain a much better understanding of how CDS market players are reacting to new information. More importantly, one can gain this understanding in the early stages of credit deterioration, or optimally, prior to it. As we will demonstrate with a few examples below, levels of market activity can be much more reactive to new information than price levels and can provide a leading indicator of future price movements.

Understanding the market activity can help flag areas of market concern to risk managers or traders before significant shifts in price movement occur. The market activity can be tracked and monitored using reliable and transparently collected data. Alert signals can be produced bringing an affected counterparty’s name to the attention of the risk managers. Such alert signals, if processed on a timely basis, can improve an institution’s ability to trigger ad-hoc assessment of a particular counterparty or credit, and potentially adjust their exposure before the potential credit deterioration.

Monitoring CDS market activity
As we mentioned earlier, it is challenging to obtain a timely source of CDS transactional history, containing both price and liquidity information. When trying to track CDS market’s trading activity, we must therefore rely on data that is closely related to trading patterns. The measures presented in this article, are based on patterns of information flow between CDS market makers and investors. We suggest that by looking not only at what prices are being quoted, but by looking at how often, and to how many counterparties a given credit is being quoted, one can effectively monitor levels of market interest in a particular entity.

In order to analyse the CDS market quoting patterns we use market activity indicators sourced from our dataset. The remainder of this article demonstrates how CMA’s CDS price and market activity indicators can enhance market monitoring processes as well as active counterparty credit risk management.

CMA market activity indicators
The dataset is collected directly from the trading desks of buy-side CDS market participants. Based on this market data, CMA produces intraday and end-of-day consensus pricing for the entire liquid CDS market.

The model creates not only reliable CDS prices, but also indicators of market activity. Market activity can be measured in terms of how often and how many market makers send indicative quotes to buy-side clients or how many of clients see a minimum threshold of quoting activity in a given reference entity. These market activity indicators are published daily as an integral part of CMA’s CDS pricing information service.

Quite often these market activity indicators reveal more about the market than CDS price levels themselves. Our market activity indicators reflect various dimensions of market makers’ quoting patterns. We assume market makers will provide indicative quotes more frequently and to a wider range of clients when facing increased trading activity in a particular reference entity. Even if the consensus price of CDS protection remains generally unaffected, increased quoting activity may signal that a larger segment of the market has begun to take a view on, or an interest, in the underlying credit.

Let’s take a closer look at two examples of a quoting pattern that can be used to draw inferences about a specific reference entity or a sector of reference entities. These examples should demonstrate how significant changes in quoting patterns could have raised institutions’ attention to possible issues with the reference credit. We suggest that a timely analysis of what has triggered CDS market interest in those names, would have improved the overall efficiency of the counterparty credit assessment process.

Greece case study
In order to provide a contemporary example, we have chosen to analyse the CDS data on Greek sovereign credit. For some years Greece has experienced severe difficulties managing its fiscal position, resulting in a massive level of national debt. Most recently there have been concerns regarding Greece’s ability to refinance itself in public markets and concerns that it would need to seek the support of either the European Union or the International Monetary Fund. Intense media scrutiny of Greece’s fiscal position started in early December as the price of Greek government bonds began plunging. As a result, many institutional investors exposed to Greek sovereign credit suffered substantial losses.

CMA1

In Fig.1 we can see how the protection price for Greece (black dotted line) started gradually rising in early November, slowly reacting to the first news of the EU’s concerns about Greece’s debt problems. However, a much clearer picture about CDS market reaction over the period of November is provided by the QF indicator (solid red line). We can see that QF for contracts on Greece rose fivefold through the nine days (November 10th-19th) after the first news on the EU’s concerns were made public. The QF rose from 477 quotes a day, to an enormous 1529 quotes a day. In Figure 2 we can see that through this period quoting activity increased around Germany as well, as it had indeed for most of Eurozone countries. However, Greece saw the biggest increase in QF levels by a significant margin, followed by Portugal (and Spain – not plotted). Notice that QF levels on countries outside Eurozone, such as Poland, remained unaffected.

CMA2

The QF figures suggest that substantial market interest in Greece and other Eurozone countries developed during the week prior to November 19. Over the two or three weeks following November 19, we can observe further widening of the basis between Greece and Germany CDS spreads in Fig.3. The period was concluded with a substantial widening of the bond yield basis in early December.

CMA3

We can conclude that the QF figures provided an early and strong signal on the changing market perception of Eurozone credits, particularly Greece, as early as on November 19. This change of perception was most clearly seen around countries with high debt to GDP ratios. However, it took almost three weeks until the CDS spread for Greece broke through the 200 bps level and started rocketing towards 400. During the three weeks leading up to December 9, any investor monitoring signals from the CDS market could have taken the opportunity to assess the magnitude of the issues that Greece was starting to face, draw relevant conclusions, and apply active measures to mitigate the increasing risk of credit deterioration.

Conclusion
We have sought to analyse the new challenges for active risk management that have been exposed by recent market turmoil. These challenges include finding a more dynamic way of monitoring and responding to changes in counterparty credit. We have proposed that previously unavailable information on CDS markets can provide reasonable support for those needs.

In order to track CDS market activity in individual credits, we chose to analyse CDS market quoting patterns using data sourced from a unique dataset. Through a combination of CDS price data and market activity indicators, risk managers can develop a more complete risk profile of specific reference entities.

Exclusive data can help identify changing market interest in a given credit by highlighting changes in CDS quoting patterns, before these changes translate into price movements. The example of Greece (and in another CMA case study Lehman Brothers) demonstrate how these market activity indicator changes provided additional signals on those entities before a significant price shift took place.