Credit Default Swap Succession
~ not delivering any success ~
Guidelines for
assessing and understanding succession issues in Credit Default Swaps around corporate events
by
Credit Default Swap Succession
~ not delivering any success ~
Guidelines for
assessing and understanding succession issues in Credit Default Swaps around corporate events
London, September 2006
Author:
Floris
F.
C.
Schorer
Student
nr:
1229923
Email:
florisschorer@hotmail.com
ING Financial Markets
Strategic Trading Platform London
Supervisor:
Drs. W.P.F. Appel
University of Groningen
Faculty of Management and Organisation
1
stsupervisor:
Dr. H. Gonenc
2
ndsupervisor:
Dr. J. H. von Eije
“To succeed… you need to find something to hold on to”
(Tony Dorsett)
Preface
The report you are about to read is the hard evidence of a six-month internship started in March 2006 at the ING Strategic Trading Platform in London. Besides that, it also marks the end of my Industrial Engineering and Management Science studies at the University of Groningen in the Netherlands. I am very grateful that I was given the opportunity to complete my studies and master thesis in such an exciting, challenging and educational environment as STP London. There are various people I owe many thanks to for giving me that opportunity.
First of all I would like to thank my colleagues at STP fixed income; Jackie, Jeff and especially Willem Appel, my supervisor and head of STP fixed income, for challenging me intellectually, introducing me to the ins and outs of the capital markets and showing me around the square mile. I also consider Willem’s contributions and feedback to my research as extremely valuable to the end result of this report. He gave me a lot of freedom in executing my research and gave me confidence in my ability to finish this report successfully by extending my stay in London.
Secondly, I would like to thank my main supervisor at the University of Groningen, Dr. Gonenc for guiding me through the process of writing a master thesis from the scientific point-of-view. I acknowledge his flexibility and involvement which shortened the distance between Groningen and London and resulted in quick communication and valuable feedback. I also owe thanks to Dr. von Eije, my second supervisor at university, for his time and comments and especially for making me enthusiastic about the world of finance in the first place.
As the completion of this thesis marks the end of a very special era, I want to use this opportunity to shortly look back at those years as well; a period in which I have undergone an enormous personal development both socially and intellectually. There are numerous people I should thank for this, but in particular my parents who gave me the opportunity to enjoy these years and Danielle, my brothers and my friends for filling in those years in a way I will always remember.
Floris Schorer
Suggested Reading Guide
This thesis is based on scientific research but its findings are focused on the practical side of the topic. Therefore it qualifies as applied research which suggests that its audience consists of two main groups; readers with a scientific interest and readers involved in the credit default swap market. Readers with a scientific interest might be more concerned with the methodology and structure of the research, whereas readers involved in the credit default swap market would be more attracted to the implications of the findings. In order to meet these various interests in the best possible way, this suggested reading guide is drawn up.
Scientific Readers
The research is structured from a scientific perspective. Therefore, readers with a scientific interest can simply follow the order of the report. If one is not familiar with the topic of this research, it is advised that chapter 1 is given extra attention. It might also be useful to read some additional literature on credit default swaps to start with a knowledgeable background. Chapters 2 and 3 will be the most interesting for this type of reader as these chapters discuss the methodology and the literature on the topic. Later on in the research, bits and pieces of the methodology will return on several occasions mainly by means of the conceptual model.
Readers Involved in the Credit Default Swap Market
Readers interested in the findings related to the credit default swap market are generally less interested in the research methodology and the scientific justification. If this is the case, it is advised that they use a different route through this report than the order in which it is set up. For the first reading it is recommended to stick to the following order when skimming this thesis:
- Executive Summary in combination with Exhibit 48 on page 94 - Succession Events (1.3.6)
- Research Objective, Main Research Question and Limitations of the Research (2.2, 2.3 and 2.5)
- If not familiar with the case studies, it is advised to read at least the summaries of the case studies (Chapter 4)
- Generalisation except for Model Adjustments (Chapter 6 except 6.4) - Guidelines for involved parties (Chapter 7)
For any subsequent and more elaborate readings it is recommended to follow the standard order of the report.
Executive Summary
Motivation
This research focuses on the Credit Default Swap (CDS), a fairly new financial derivative that is designed to transfer the credit risk of an underlying asset to a third party in order to protect for a default. Recent increases in corporate actions, such as M&A, have exposed shortcomings in the way the CDS and its regulation is structured by the International Swaps and Derivatives Association (ISDA): In some corporate events a CDS fails to cover the credit risk it was intended to cover in the first place. In such situations, the underlying asset from which the CDS was derived no longer qualifies as deliverable to execute the swap following a default. This leads to asymmetric behaviour of the price of a CDS; as a result of a corporate event the CDS loses its connection to the underlying asset it was designed around, is “orphaned” and thus detaches from the credit risk it was intended to protect. The price of an orphaned CDS no longer reflects the credit risk of the business it was designed around.
As the CDS is an important investment instrument to, among others, the fixed income team of the ING Strategic Trading Platform (STP), an in-depth analysis of this issue can reduce the volatility of earnings on STP’s CDS positions and enhance their understanding of the CDS. Moreover, it could lead to more efficient pricing of the CDS in general and possibly credit spreads in general.
Objective
Determine the relationship between ISDA succession language and credit default swap spreads in order to clarify asymmetric spread behaviour around corporate events and hence reduce the risk in trading credit default swaps for STP.
Methodology
The issues around succession of a CDS and deliverability into a CDS have only arisen very recently. Data and cases on the issue are limited. Prior research related to the CDS has always focused on the economic determinants of CDS spreads. Therefore, this research is set up in an explorative way; it focuses on relationships between the various aspects of the problem. As a result the research is qualitative by nature and its findings are based on various case studies that illustrate the diversity of issues and cover every part of the problem.
The six companies used for the case studies are: Sainsbury, Rentokil, Cendant, Alltel, ProSieben and ITV. It is determined whether or not asymmetric behaviour of the CDS spreads occurs in each case by comparing changes in corporate variables to changes in CDS spreads. The variables or determinants are selected based on findings of prior research of, among others, Cossin & Hricko (2001) and Hull et al. (2004). The three determinants used in this research are credit ratings, equity volatility and general
macro-economic variables. Once it has been established that the behaviour of the CDS spreads can be classified as asymmetric in comparison to the underlying credit risk, the cause for this behaviour is examined.
With the findings of the case studies a generalisation is made for succession issues of a CDS in relation to corporate event categories by means of two models. These two models provide guidelines for the way a CDS should be treated around corporate events.
Findings
The case studies all show a noticeable degree of asymmetric behaviour of the CDS spreads which is (indirectly) caused by ISDA succession language. In five of the six cases the issue is related to (a fear of) the CDS ending up with no deliverable obligations. The accompanying spreads show a considerable tightening in all of these cases. It is found that the occurrence of the issues in the cases is related to at least one of three critical event categories: (i) a restructuring of the company, (ii) a refinancing of the company (where refinancing also includes the increase of leverage) and (iii) the withdrawal or insertion of a guarantee. Moreover, a company’s characteristics facilitate identification of those companies that are relatively more likely to be involved in these types of event categories. These findings led to the design of the Driver Model on page 69. The model lists empirically identified company characteristics with a propensity to lead to the three critical event categories; early indicators. These company characteristics are grouped by five different categories: Structure (which also includes capital structure), Performance, Stakeholders, Competitors and External Variables.
Embroidering on the Driver Model, the link between these three event categories and succession issues is made. It results in the Succession Model which determines the likelihood of succession issues by using the characteristics of the upcoming corporate event. The higher the number of the cell the event would qualify for, the higher the likelihood of asymmetric behaviour of CDS spreads because of succession issues. Guarantee-related issues may add to the risk as well. However, these are too case specific to be incorporated in the Succession Model.
The names of the cells illustrate typical events that would qualify for that cell e.g. a break-up would typically be an event which focuses on a high-impact restructuring and any refinancing would be a side-effect of it. A break-up would therefore qualify for cell-3. The likelihood of succession issues in cell-3 events is fairly high. refinancing impact re st ru ctu rin g impact high lo w low high Break-up / Merger Stand-alone LBO Low Risk Cell-1 Cell-2 Cell-3 Cell-4
It is concluded that ISDA succession language has influence on CDS spreads around corporate events, particularly if the event involves restructuring, refinancing and/or guarantee risk. The influence is caused by shortcomings in the legal structure of the CDS. The market uses an economic approach to the CDS, i.e. it assumes that the CDS pricing will follow the core credit risk, irrespective of any corporate event. However, the ISDA structure fails to capture the economic approach used by the market. That shortcoming is causing the influence. Events can be categorised by the likelihood of this influence actually surfacing which helps investors like ING STP understand and anticipate the risk in trading a CDS.
Guidelines
Based on these findings, guidelines are drawn up for the three main parties involved in the CDS market.
The first involved party, the ISDA, is the party that has the most influence in solving the problems around corporate events as it provides market standards of the CDS; it has the ability and power to amend the 2003 ISDA Credit Derivatives Definitions. It also caused the problems by failing to capture the economic approach to CDS in its definitions. If the ISDA decides to change the definitions the focus should be on the inclusion of debt-replacements like a refinancing in the definition of a Succession Event. In addition, the ISDA should create an option to recognise certain events as a Succession Event even though a guarantee would block this according to the definitions. These measures would address the main issues of refinancing and guarantees that were identified in the research. Secondly, corporate borrowers, i.e. the companies that are the subject of a CDS contract, should use their best efforts to maintain a stable CDS market and reduce succession problems were possible as this will enhance their relationship with the capital markets. A step further would be to include such a best-efforts undertaking in a clause in the company’s bond documents.
Thirdly, market participants are able to apply strategies to anticipate succession issues. To assist in these strategies, three different phases are identified during the course of events that involve a high-impact restructuring: Rumour, Announcement and Completion. In the Rumour phase CDS spreads are priced according to the economic approach to a CDS; prices represent the credit risk of the operations of a company and little notice is given to whether or not obligations are deliverable into the CDS. In the Announcement phase, the approach to price a CDS switches unpredictably between the economic and legal approach. This is a highly speculative phase and only investors with a high appetite for risk should take positions in a CDS in this phase. In the Completion phase, the legal approach prevails over the economic approach and CDS spreads are based on the ISDA definitions in combination with the details of the event. Investors should have excellent knowledge of the legal and structural aspects of a CDS if they invest during this phase.
Recommendation for further research
Further research elaborating on this subject should include a quantification of succession issues. This is possible once more cases involving succession issues have been identified and more extensive data is available. Also research focusing on the legal side of succession issues will enhance solving the problems and could assist the ISDA in amending the 2003 Credit Derivative Definitions.
Table of Contents
Preface ...iii
Suggested Reading Guide...iv
Executive Summary ...v
Table of Contents ...ix
1 Introduction... 1
1.1 Motivation of the Research ... 2
1.2 ING Financial Markets – STP ... 3
1.3 The Credit Default Swap... 4
1.3.1 Basics...4 1.3.2 History...6 1.3.3 Pricing ...7 1.3.4 Market ...7 1.3.5 Credit Events ...10 1.3.6 Succession Events ...11 2 Methodology ... 14 2.1 Problem Definition ... 14 2.2 Research Objective... 15
2.3 Main Research Question ... 15
2.3.1 Sub-questions...15
2.4 Conceptual Model ... 17
2.5 Limitations of the Research ... 18
2.6 Nature of the Research ... 19
3 CDS Spread Determinants... 21 3.1 Company-Specific Determinants... 23 3.1.1 Capital Structure ...23 3.1.2 Firm Value...23 3.2 Macro-economic Determinants ... 24 3.3 Relevant Determinants ... 25 4 Case Studies ... 26 4.1 Sainsbury... 27 4.1.1 Issue ...27 4.1.2 Effect...28 4.1.3 Analysis ...28 4.1.4 Summary ...32
4.2 Rentokil Initial plc ... 33 4.2.1 Issue ...33 4.2.2 Effect...34 4.2.3 Analysis ...34 4.2.4 Summary ...37 4.3 Cendant Corporation... 38 4.3.1 Issue ...38 4.3.2 Effect...39 4.3.3 Analysis ...39 4.3.4 Summary ...42 4.4 Alltel Corporation ... 43 4.4.1 Issue ...43 4.4.2 Effect...45 4.4.3 Analysis ...45 4.4.4 Summary ...48
4.5 ProSiebenSat.1 Media AG / Axel Springer ... 49
4.5.1 Issue ...49
4.5.2 Effect...49
4.5.3 Analysis ...50
4.5.4 Summary ...53
4.6 ITV plc / Carlton Communications plc ... 54
4.6.1 Issue ...54 4.6.2 Effect...54 4.6.3 Analysis ...56 4.6.4 Summary ...58 4.7 Conclusion... 58 5 Analysis ... 59 5.1 Company Characteristics ... 60
5.1.1 Structure / Capital Structure ...60
5.1.2 Performance ...60 5.1.3 Stakeholders...61 5.1.4 Competitors...61 5.1.5 External Variables...61 5.2 Corporate Event ... 62 5.3 Spread Behaviour... 63 5.4 Classification ... 65 6 Generalisation... 68 6.1 Drivers... 68 6.1.1 Restructuring ...68 6.1.2 Refinancing ...70 6.1.3 Guarantees...71
6.1.4 Limits of the Driver Model ...72
6.2 Events ... 73
6.2.1 Scope of Various Events ...74
6.2.2 Succession Issues per Cell ...76
6.2.3 Drivers per Cell ...77
6.3 Assessing Succession Issues... 78
6.4 Model Adjustments ... 80
7 Guidelines for Involved Parties... 83
7.1 ISDA ... 83
7.1.1 Definition of Succession Event ...83
7.1.2 Guarantor ...84
7.1.3 Non-deliverable Insurance ...84
7.2 Corporate Borrowers... 86
7.2.1 Providing Information...86
7.2.2 Best Efforts Clause...87
7.3 CDS Investors... 88
7.3.1 Low Risk & Stand-alone Events (cell-1 & cell-2)...88
7.3.2 Break-up / Merger & LBO Events (cell-3 & cell-4) ...88
7.3.3 Single Name CDS Options ...92
7.3.4 Overview...93
7.4 Conclusion... 95
8 Conclusions and Recommendations ... 96
8.1 Achievement of the Research Objective... 96
8.2 Recommendations for further research ...100
References ... 102
Books, Articles and Research ...102
Websites and other ...106
List of Exhibits... 107
Appendices ... 110
I. Restructuring Clauses in a CDS ...110
II. Reuters release on GUS demerger ...112
III. Sainsbury...114
a) Company description... 114
b) Financials ... 116
c) S&P release on Sainsbury ... 117
IV. Rentokil ...118
a) Company description... 118
b) Financials ... 119
c) S&P releases on Rentokil ... 120
V. Cendant ...123
a) Company description... 123
b) Financials ... 124
c) Credit rating releases on Cendant ... 125
VI. Alltel...130
a) Company description... 130
b) Financials ... 130
c) Credit rating releases on Alltel ... 132
VII. ProSiebenSat.1 / Axel Springer ...136
a) Company description, ProSiebenSat.1... 136
b) Company description, Axel Springer ... 136
d) Fitch releases on ProSieben... 140
VIII. ITV...143
a) Company ... 143
b) Financials ... 144
c) Fitch releases on ITV... 145
1 Introduction
The two words ‘success’ and ‘succession’ show a noteworthy resemblance. However, there seem to be some differences illustrated by the Oxford Dictionary:
Success ● noun 1 the accomplishment of an aim or purpose.
2 the attainment of fame, wealth, or social status. 3 a
person or thing that achieves success.
-ion ● suffix 1 forming nouns denoting verbal action or an
instance of this: communion. 2 denoting a resulting state or product: oblivion.
Combining the two into ‘succession’ would lead to a sense close to the following: ‘A
resulting state of the accomplishment of an aim or purpose’. However, the Oxford
dictionary also distinguishes ‘succession’ as a word on its own:
Succession ● noun 1 a number of people or things following
one after the other. 2 the action, process, or right of inheriting an office, title, etc. 3 Ecology the process by which a plant community successively gives way to another until stability is reached.
The gap between the meaning of ‘succession’ when derived from ‘success’ and the original meaning of ‘succession’ demonstrated by the much respected dictionary seems to be evocative for the same two terms in relation to the credit derivatives market; ‘success’ and ‘succession’ don’t appear to go together in this market.
This thesis, written at ING as part of the concluding phase of the author’s studies at the University of Groningen, The Netherlands, will aim to reduce this gap with regard to the credit derivatives market. Hopefully, parties involved in this market are soon able to add another definition of succession to their own little Credit Default Swap (CDS) dictionary ● 4 A state resulting from the accomplishment of a purpose: a successful state.
The background of this research will be provided in this first chapter. First the motivation for this research will be addressed. After that a brief description of the ING department where this research was initiated will follow. Then the fundamentals of the credit default swap will be discussed along with its history, its market and its (for this research) most important characteristics. After this chapter the reader should have a reasonable understanding of the credit default swap. However, the author assumes that any reader of this research has some knowledge of financial markets.
1.1
Motivation of the Research
Enormous growth in volume and importance of the CDS market has led to the exposure of shortcomings in the structure of the CDS. These shortcomings are addressed in this thesis. The research was initiated because of two reasons related to these shortcomings.
The first and foremost reason is the growing disorder in the CDS market caused by these shortcomings. The disorder leads to incorrect pricing of credit default swaps which in turn could lead to investors avoiding the market and a erroneous perception of the credit risk of companies; a general increase in inefficiency of financial markets. To illustrate the significance of the problem to the CDS market, the result of a survey performed by Merrill Lynch is shown below.
Exhibit 1 - Survey illustrating the significance of succession issues1
The survey shows that almost one in every two investors in the CDS market is less comfortable in investing in CDS due to these shortcomings.
The second reason is ING’s involvement in the CDS market. The Strategic Trading Platform, the department of ING Financial Markets where this research was executed, relies for its earnings partly on investments in credit default swaps. As these issues decrease the predictability of the market, they increase the volatility on ING’s earnings; an effect that is not wanted.
In paragraph 2.1 the issues are discussed somewhat more which should also enhance the understanding of the relevance of this research.
1 Source: Jónsson et al (2006), the survey was published at 8 May 2006.
Are you more or less comfortable than at the start of the year in shorting credits via CDS due to potential
CDS succession issues? no change 33% more 21% less 46%
1.2
ING Financial Markets – STP
The Strategic Trading Platform (STP) is the proprietary trading division of ING Financial Markets. In proprietary trading a firm trades for its own direct gain instead of earning commission by trading on behalf of a client. STP has two main departments, one in London and one in New York. The primary objective of the division is to make consistent economic value added profit through proprietary risk taking. The division is organised around the various asset classes in which investments are made, namely:
• Equity • Fixed Income • Special Situations • Structured Products
At the fixed income team, profits are made through investments in bonds and credit default swaps. However, most investments are focused on bonds as the CDS market is relatively new. The investments are subject to a limitation on the risk exposure and the capital requirements of the taken positions. Therefore trades with low risk exposure and unfunded trades can be valuable to the department. Contrary to bonds, credit default swaps are unfunded and require less capital than bonds. These instruments are therefore of interest to STP. Besides that, a CDS provides an easy way to effectively short credit risk (see section 1.3.4) and, due to enormous growth, its market today is bigger than the bond market in terms of notional outstanding.
Considering these developments there are incentives for STP to fully understand CDS fundamentals and technicals and then possibly increase the use of the CDS in its trading strategies. An increased understanding of CDS behaviour will reduce perceived risk and increase the profit potential of CDS trades.
1.3
The Credit Default Swap
This section addresses the main characteristics of the credit default swap. Consecutively the following topics will be discussed: the basic working of the credit default swap and its definitions, the history, the way a credit default swap is priced, its market, the events triggering the swap and the for this thesis most important characteristic; succession events.
1.3.1 Basics
A credit default swap is an over-the-counter derivative2 which gives the participants the opportunity to transfer credit risk without transferring the underlying credit asset. The buyer of the CDS protects himself against the loss of principal due to a credit event in the underlying asset (see the section Credit events for the definition of a credit event). It means that the buyer insures himself against the loss of payments on money he lent to a third party. So although the buyer of a CDS lent money to a company, he now doesn’t have to worry anymore about getting his money back as this is insured by the CDS. In exchange for this protection the buyer pays a periodic premium (CDS spreads), quoted as basis points (bps) per annum over the contract’s principal (the insured amount), to the protection seller until the contract matures or a credit event occurs, whichever is earlier. The seller of the CDS is therefore the party that insures the payments; the insurer. In exchange for the premium the seller agrees to take over the risk of the obligation. When a credit event occurs i.e. when the reference entity (the company the CDS is based on) can not meet its obligations, the protection buyer delivers a deliverable obligation (a credit asset i.e. bond or loan or other obligation that fulfils the requirements of the contract) to the seller. In return the seller pays the par value of that obligation to the buyer. As a result, the seller now owns the obligation which has decreased in value because of the credit event, whereas the buyer has his borrowed money back. The contract has now been settled and terminated. This way of settling is known as physical settlement. It is also possible (depending on the agreements made when the contract was initiated) to use cash settlement. In cash settlement, the seller doesn’t receive the actual obligation. The seller only pays the difference between the par value and the market value of the reference obligation to the buyer. For example if the value of a bond (determined by a party defined in the contract, but usually the seller) after the credit event is 40% (prices of bonds are quoted in percentages of the notional amount), the seller just pays 60% of the notional amount to the buyer. With cash settlement, the buyer therefore doesn’t even have to own the obligation to settle the contract after a credit event. This way of settling encourages trading risk of a company’s obligations without dealing with the obligations the risk is based on.
By using a CDS one is able to diversify portfolio risks without diversifying the portfolio itself. This obviously can be useful for financial institutions having to meet their capital
adequacy requirements and at the same time wanting to keep a good relationship with clients. However the CDS market can be attractive for speculation as well and therefore attracts all kinds of financial investors (see section 1.3.4).
Exhibit 2 - CDS basics
Definitions
Credit derivatives: Comprise credit default swaps, credit-linked notes and portfolio swaps. As CDS contracts account for over 80% of the credit derivatives market, this research will focus on the CDS and use the terms CDS and credit derivative interchangeably.
Credit risk: Comprises the possibility of a loss occurring due to the financial failure to meet contractual debt obligations.
Credit asset: Is the extension of credit in some form, for example a loan, bond or financial lease contract. In a CDS contract it is defined by a reference obligation of a specified reference entity.
Deliverable obligation:3Comprises any obligation that can be delivered to settle the contract when a CDS is triggered. Besides the reference obligation specified in the contract, this includes any obligation of the reference entity that is not subordinated to the reference obligation provided that its denoted currency is one of the following: Euro, US
3 For the full legal definition, see 2003 ISDA Credit Derivatives Definitions
CDS / protection
buyer
CDS / protection
seller
Premium
CDS / protection
buyer
CDS / protection
seller
Par amount - Market value*
Cash settlement
CDS / protection
buyer
CDS / protection
seller
Deliverable obligation
Par amount
Physical settlement
Settlement after a credit event
Before a credit event
* Market value is often not used here. More
common is a negotiated cash settlement price
Dollar, British Pound, Japanese Yen, Canadian Dollar and Swiss Franc. Additional restrictions can be made if the triggering credit event is a restructuring (see credit events).
Reference entity:4 The precise name of the legal entity on which the CDS contract provides protection.
Reference obligation: The particular obligation as it is specified in the CDS contract. Although this is only one obligation, more obligations are usually deliverable into the CDS contract.
1.3.2 History
In the early 1990s, banks forged CDS contracts by purchasing protection from insurers in order to manage their exposure to corporate loans on their books. International rules promulgated under the 1988 Basel I Accord further fuelled interest in the CDS by requiring banks to set aside a greater percentage of their capital against outstanding loans. Banks, anxious to transfer credit risk to entities that weren’t subject to the same capital reserve requirements, saw the obvious benefits of the CDS, and understood that they could transfer risk while still retaining the ownership and profits of the loan, as well as maintaining relationships with their clients.5
The first CDS contracts evolved and in 1999 the International Swaps and Derivatives Association (ISDA) issued definition on credit derivatives which were altered and extended in 2003. These definitions are now used as the market standard for CDS contracts. As hedge funds and other institutional investors entered the CDS market, contracts became standardised and markets more liquid. Moreover, recent research by the IMF provides evidence that changes in CDS spreads lead changes in bond spreads in the short term nowadays.6 This is a result of the enormous growth undergone by the credit derivatives markets in recent years as is illustrated by Exhibit 3.
Exhibit 3 - The Bond market compared to the CDS market7 ~ notional amounts outstanding (trillions of US dollars) ~
4 For the full legal definition, see 2003 ISDA Credit Derivatives Definitions 5 www.fenews.com/fen48/one_time_articles/credit-deriv/credit-derivatives.html 6 International Monetary Fund (2006)
1.3.3 Pricing
CDS spreads reflect the expected cost of providing protection. The present value of the premium payments for being protected should therefore equal the present value of the expected losses for the protection seller. This expected loss is calculated from the probability of a default of the reference entity happening and the expected loss when such a default occurs, expressed by the recovery rate (see the equation below). For example, if a bond investor expects to receive 40% of his investment back following an event of default, the recovery rate would be 0.4. In other words, the seller of the CDS has to cover that 60% loss. So if the seller expects an event of default on this bond in three years time, he wants the premium he is getting paid to be at least equal to that 60% he has to pay in three years. The mathematics behind calculating the probability of a default is beyond the scope of this research. For a detailed explanation on the calculations of these probabilities see Hull and White (2000). However, the variables influencing this probability and thus CDS spreads will be addressed later on in this research.
The equation below concretizes the pricing of the CDS. The first part of the equation reflects the periodic premium payments. The 0.000025 factor is to adjust for the fact that spreads are quoted in basis points (factor 10,000) per year while they are paid per quarter (factor 4).
( )
( )
(
)
( )
( )
(quarter) period obligation protected the of rate recovery p quarter in entity reference for the default of y probabilit CDS the of protected) (amount amount notional s basispoint in spread contract CDS points percentage in rate discount quarterly contract CDS the of maturity until quarters of number 1 1 1 * * 1 1 * * 000025 . 0 = = = = = = =∑
∑
= = + − = = +=
p R DF N S r q where p q p p r R N DF q p p r N S p p pρ
ρ1.3.4 Market
As the CDS is an over-the-counter instrument, there is no exchange for the CDS. However, the market for the CDS is made by brokers providing bid-offer prices. Especially the market for CDS contracts with five year maturity is very liquid nowadays. This liquidity was encouraged by the standardisation of the contract. The most liquid CDS names are united in indices such as the iTraxx-index in Europe and the CDX-index in North America. The
different indices vary in number of constituents from 30 reference entities to 125 reference entities and provide a general view on the credit market.
Participants in the CDS market are:8,9 - Bank portfolio managers - Bank proprietary desks - Hedge funds
- Insurance companies and pension funds - Corporate treasury
- Credit correlation trading desks
Most of these participants invest in the CDS for capital gains like STP does. They usually don’t invest in the CDS because they own obligations and are afraid of credit events, but they invest because they anticipate changes in credit risks. For them the CDS provides an easy way to effectively short credit risk or to be effectively long credit risk without the needed funds. Exhibit 4 shows the working of that.
Exhibit 4 - Relationship between bond spreads and CDS spreads10
In Exhibit 4 an investor buys an obligation (for example a bond) at par value and he buys a CDS for protection of the total amount of that par value, hence this investor is not exposed to any risk (hereby ignoring the risk that the protection seller can not meet his protection obligations in case of an event of default i.e. counterparty risk). As this investor
8 Mark-it (2005)
9 For more information on market participants and their reasons, see Risk Waters Group (2001) 10 Based on Mahadevan, Polanskyj et al (2006)
RISK
Obligation with fixed coupon
Asset Swap
RISK FREE INVESTOR
Obligation buyer (risk buyer)
CDS buyer (buys protection against risk)
financing
RISK TAKER
CDS seller
(sells protection and assumes risk)
Par value
Libor
Coupon
Par value
Credit risk
Par value
Libor and asset
swap (credit ) spread
Credit risk
CDS premium (spread)
Interest rate risk
and credit risk
doesn’t have any risk and there is no such thing as free money (in a perfect market), his investment should be cash-neutral.
The investor buys the fixed coupon bond at par financed at Libor11 (the London InterBank Offered Rate). In return he receives interest rate risk (because he receives a fixed coupon financed at the varying rate of Libor), credit risk (the risk of a default on the bond) and of course the coupon. It is possible for the investor to eliminate his interest rate risk by means of an asset swap. This swap effectively changes the fixed coupon rate into a floating rate quoted in spread over Libor. As he also finances the trade at Libor, he is now no longer exposed to interest rate risk. The investor is left only with exposure to the credit risk. In return for this exposure he receives Libor + the asset swap spread. Again, his investment was also financed with Libor so effectively he assumes the credit risk in return for the asset swap spread.
When he buys the CDS, he pays a premium to sell the credit risk. Logically, this premium should be equal to the asset swap spread he received for assuming the credit risk. Therefore, bond spreads are equal to CDS spreads with equal maturities in a perfect market and the bond trading at par.12
This characteristic shows why investors use a CDS to effectively short a bond; by buying a CDS, the investor is short the credit risk just like his position if he were to short a bond. By buying the CDS though, he doesn’t need to go trough the trouble of borrowing a bond. This works the other way around as well; by selling a CDS, one is effectively long a bond but without the funding.
The CDS market is therefore used to invest just like the bond market and because of the favourable characteristics of the CDS market, this market has outgrown the bond market.
As a CDS is a contract, taking profits is not as straightforward as with shares or bonds. There are two ways to step out of the contract.13
First, it is possible to tear up the contract upon agreement with the other party. This way, the contract is terminated. If the market spread at the time of the termination is different from the contract spread, the present value of this difference is paid e.g. if an investor bought a CDS at 40 bps and at the time of the termination the market value of this contract is 60 bps, upon termination of the contract he receives the present value of those 20 bps difference from the protection seller. The protection seller effectively buys back the contract at 60 bps.
The second way of taking profit is passing the contract to another party, which is called assignment. Unlike the first process, in this process the contract still exists after the transaction. A buyer of a CDS effectively sells his contract to a third party who is willing to buy that CDS. Again the difference between the market and the contract spread is paid by calculating the present value of the payments due until the contract ends.
11 This assumes that an investor can fund at LIBOR
12 There can be slight differences because of, among others, options in the bond document influencing the modified duration of a bond e.g. call options. However, this is beyond the scope of this thesis and not relevant for the succession issues. For a more elaborate discussion on these differences, see Risk Waters Group (2001). This topic is also addressed by Benkert (2004).
13 The ways of taking profit on a CDS contract are somewhat simplified as the purpose of this explanation is not to exactly calculate those profits.
To summarise, buyers of a CDS enjoy protection against default and are therefore effectively short the credit risk. Hence, they expect the credit risk to increase together with the spreads on their CDS. Sellers of a CDS are exposed to events of default and are effectively long the credit risk. Hence, they expect that risk to decrease and therefore the spreads on the CDS to tighten because the probability of a default would be lower.
1.3.5 Credit
Events
A credit event triggers the CDS, meaning that settlement of the contract will occur, either by means of physical settlement or cash settlement. A credit event is basically what a CDS buyer originally protects himself against. In recent years however, credit events are rare as the credit market enjoys a period of statistically low default occurrence.
The events constituting a credit event are defined in each separate CDS contract. In a standard CDS contract however, bankruptcy and failure to pay are always included as credit events. According to O’Kane, Pedersen and Turnbull (2003) these are called hard credit events because after such an event occurs bonds ranked pari passu (equal) should trade around the same price. That is because all bonds will be accelerated at the same time in such an event. Any obligation of the reference entity that is not subordinated to the reference obligation can be delivered into the contract (see definitions).
Restructuring of a bond is also considered a credit event in certain situations. However, this event is different from the first two and is known as a soft credit event. This is because after such an event, obligations don’t become immediately due and will continue to trade at a price differential. As a result protection buyers can choose which deliverable obligation they will deliver to settle the contract and thus have the option to deliver the cheapest obligation. This is what happened after a restructuring of loans at Conseco (US insurance company, September 2000); buyers of protection who also held short term loans claimed a credit event had occurred after the maturity of those loans was extended. To settle the CDS contract however, they delivered long-term obligations (instead of the short term loans) which traded at significant lower prices than their short term loans. This way, they were able to make money on this credit event at the expense of protection sellers.14 According to the 2003 ISDA definitions a debt obligation is considered ‘restructured’ if one of the following conditions are fulfilled:
• The interest rate on the obligation is reduced • The principal or premium is reduced
• The maturity is extended
• The priority ranking of payments is changed
• The currency or composition of the payment of principal or interest is changed
In any case of these five different restructurings of bonds, a protection buyer or seller can claim an event of default. To provide the possibility to eliminate or at least reduce the ‘Cheapest-to-Deliver’ opportunity as it was experienced at the Conseco case, the ISDA came up with four different clauses regarding restructuring. More information on these clauses is provided in Appendix I.
1.3.6 Succession
Events
In this section the concept of a succession event and succession language will be explained. As succession events are the core of this thesis, this section is very important. The reader should bear in mind that a succession event is in no way a credit event; hence a succession event will not trigger the CDS to be settled. Even though corporate events often involve a restructuring, a reader should distinguish between a restructuring causing a credit event i.e. a restructuring of a bond itself and the restructuring of a company or its capital structure possibly causing a succession event.
In recent years volume of corporate actions such as mergers, acquisitions, disposals etc. has rapidly increased. While these M&A waves have occurred in the past, the last few years mark the first M&A wave which the relatively young CDS market is experiencing. It is therefore widely regarded as a test case for ISDA 2003 language; is the language constructed in the way that the CDS will still reflect the right risk after a corporate action? In most of these actions capital structures were altered, resulting in a change of deliverable obligations on the CDS. As companies may cease to exist, or their debt is exchanged following M&A activity, rules had to be defined for the implications of these actions on the CDS; both for the implications for the reference entity and for the reference obligation. These rules are known as succession language and an event where a reference entity is replaced by another is known as a succession event. The 2003 ISDA definition for a succession event is as follows:
“Succession Event” means an event such as a merger, consolidation, amalgamation, transfer of assets or liabilities, demerger, spin-off or other similar event in which one entity succeeds to the obligations of another entity, whether by operation of law or pursuant to any agreement. Notwithstanding the foregoing, “Succession Event” shall not include an event in which the holders of obligations of the Reference Entity exchange such obligations for the obligations of another entity, unless such exchange occurs in connection with a merger, consolidation, amalgamation, transfer of assets or liabilities, demerger, spin-off, or other similar event.
This language is important as it defines which reference entity the CDS can and will refer to, following a corporate event. In some cases this entity can be different from the entity any given debt obligation is transferred to. The term “succeed”, as it is used in this definition is defined by the 2003 ISDA definitions as well:
“succeed” means, with respect to a Reference Entity and its Relevant Obligations (or, as applicable, obligations), that a party other than such Reference Entity (i) assumes or becomes liable for such Relevant Obligations (or, as applicable, obligations) whether by operation of law or pursuant to any agreement or (ii) issues Bonds that are exchanged for Relevant Obligations (or, as applicable, obligations), and in either case such Reference Entity is no longer an obligor (primarily or secondarily) or guarantor with respect to such Relevant Obligations (or, as applicable, obligations).
Because of the exchange exclusion in the second part of the “Succession Event” ISDA definition, the timing of actions can be crucial to whether these events would qualify as succession events or not. If an exchange without a connection with a corporate action would occur, there will be no succession event, possibly leaving the CDS without a deliverable obligation as it is possible then that the CDS refers to a reference entity without any debt. The CDS will then become worthless as it insures debt that doesn’t exist (see Case studies for examples). A CDS without deliverables is also called an orphaned CDS; the instrument the derivative is derived from (the bond issued from an entity) ceases to exist or the “parent” of the CDS is gone.
Considering the succession language, the way in which a firm is structuring a corporate action is very important to the future of the CDS. As the activity of corporate events is growing, numerous ways to structure these events are developing and these structures get more complicated. Because of this complication it is often not clear from the announcement of a corporate action if a succession event is about to happen or not. A related issue concerns guarantees given to companies by their (newly acquired) subsidiaries. As can be seen in the “succeed” definition, guarantors of debt are excluded from succession events. This means that when company A acquires company B and A assumes all of B’s debt, a succession event occurs unless B guarantees the assumed debt. Hence the reference entity in the CDS contract in such situation will only be changed from B to A if B doesn’t still guarantee the by A newly assumed debt. This is a subtle difference but therefore with huge consequences for the CDS. It is often not clear, at the time of the announcement of a corporate event, what the structure of guarantees will be and if there’ll be guarantees at all.
When a succession event is recognised, the issue of which part of the debt is actually transferred to which company arises. This is important in determining any change in a CDS Reference Entity. In Exhibit 5 different scenarios are listed with their consequences for the CDS contracts.
Exhibit 5 - Determining Successors15
Debt assumed by New reference entity
1 company ABC assumes ≥ 75% company ABC is the successor
2 One or more entities assume 25% - 75% Contract equally divided among these entities (all successors)
3 No entity assumes more than 25% Original legal entity stays reference entity (no changes)
4 The legal (old) entity does not survive Follows the succeeding entity with the highest % of obligations
Say company X assumes 30% of a company Y’s debt (note that this is not just the deliverable obligations) and company Z assumes the other 70%. If one would have bought a CDS on Y before this transaction, the reference entity will change following this transaction. According to the ISDA definitions summarised in Exhibit 5, this CDS will be split equally between X and Z even though Z assumed 70% of the debt. This means that the notional amount initially protected for default on the deliverable obligations of Y will now be split between the deliverable obligations of X and Z.
Succession can be seen as a two-step process. The first step involves establishing whether or not an actual succession event has occurred according to ISDA definitions. If it is confirmed that a succession event has occurred, the second step consists of determining one or more of the successor entities that will become the new reference entities in the CDS contracts.
For an illustration of the relevance of succession events in the CDS market, see Appendix II-Reuters release on GUS demerger.
2 Methodology
This chapter will elaborate on the problem this thesis is based on. After a description of the problem, the research objective is formulated together with the main research question and related sub-questions. A conceptual model is designed to structure the problem. Following this, the limitations of the research are formulated and the method of research is chosen.
2.1 Problem
Definition
Credit Default Swaps were designed to transfer the credit default risk of a corporate borrower among investors. The corresponding spreads on a CDS generally trade on the basis of the credit risk of the underlying business of the reference entity. To put it more clearly, credit risk is comprised to a large degree of the overall risk profile of a company’s operations and finances. The risk is not priced according to the risk of a single reference entity in isolation as such reference entity is assumed an integral part of a company’s operations as a whole; the economic approach to credit risk and a reference entity used by the market. In CDS contracts however, obviously a legal approach is used. As the CDS is fairly new and its market has been growing exponentially, imperfections in its legal design and consequently its pricing are surfacing; conflicts occur between the market’s economic approach to risk pricing and ISDA’s legal definitions. With the problem of restructuring events and the maturity of possible deliverables only just solved by introducing different restructuring clauses (see the section Credit Events and Appendix I), it seems a new problem is arising already.
Succession language on the CDS seems to become an issue for CDS spreads as general corporate activity is rapidly increasing, largely fuelled by increasing M&A and LBO activity. With this increasing activity, numerous different ways of structuring these events are found e.g. the choice between a spin-off and a split-up. Corporate management of the companies involved in these activities is obviously less concerned (if at all) with the consequences of a deal for CDS participants than they are about strategic and operational issues. Credit Suisse provides a clarifying analogy about the confusion caused by some of the structures of events: “If you changed the name of your first child on the birth date of your second child, and named the second child with the first one’s original name, do you think there would be some confusion?”16
These corporate actions resulted in cases where succession language in ISDA Credit Derivatives Definitions confused the market on the behaviour of CDS spreads in comparison to the underlying credit risk levels. Thus, it seemed that CDS spreads didn’t represent the accurate credit risk levels; a discrepancy between a legal (ISDA) and an
economic (market) approach. Because of the content of the succession language, CDS market participants are afraid that there won’t be any deliverable obligations left at the reference entity of the CDS contract after a given corporate event. As a result, spreads tighten while the underlying credit risks of the companies don’t change or maybe even worsen.
This asymmetric behaviour increases the risk of trading the CDS as it decreases the predictability of CDS spreads around corporate events. This, in turn, can result in unforeseen losses not related to a change in credit risk. A study which can clarify this asymmetric behaviour related to succession language would therefore be valuable to STP as it will lead to more predictable CDS spreads. Moreover, it may even create a competitive advantage on anticipating spread behaviour in these situations.
2.2 Research
Objective
The research objective shows the ultimate goal of this research and its relevance. It also helps the researcher to focus on the problem. The objective is derived from the problem definition and reads as follows:
“Determine the relationship between ISDA succession language and credit default swap spreads in order to clarify asymmetric spread behaviour around corporate events and hence reduce the risk in trading credit default swaps for STP.”
2.3
Main Research Question
The main research question is deducted from the research objective and will help the researcher to achieve the objective. It makes the objective more concrete and practical. The main research question is stated as follows:
“Can asymmetric behaviour of credit default swap spreads around corporate events be explained by ISDA succession language and if so, in what way can parties involved in the CDS market take this relationship into account?”
2.3.1 Sub-questions
In order to come to a satisfying answer on the research question, the problem will be divided into consecutive sub-questions which will all contribute to the research objective. The questions each address a manageable chunk of the problem.
1. What are the known determinants of CDS spreads?
By addressing the known determinants of CDS spreads, one is able to find out if asymmetric behaviour of CDS spreads has occurred in certain situations. Such a situation will be distinguished by occurring spread behaviour that can not be explained by those determinants identified in this sub-question.
2. In what cases did corporate events cause issues around the succession of a CDS?
This question will show the variety of issues related to succession language. Six case studies will provide the relevant information. Several topics will be addressed to show how succession issues come about and to illustrate the occurrence of asymmetric behaviour in practice:
a. What were the characteristics of these events?
b. How did the CDS spreads evolve around these events?
c. Did the credit risk levels change? / Did the known determinants of the CDS change?
d. If there was any asymmetric behaviour, what caused it? 3. Are there similarities among these cases?
In order to generalise succession issues and provide a covering framework it is important to identify similarities between the case studies. These similarities will be categorised among three different attention points:
a. Similarity among company characteristics b. Similarity among corporate events c. Similarity among CDS spread behaviour
4. What are guidelines for parties involved in the CDS market when considering corporate events and accompanying succession issues?
The last sub-question is aimed at providing guidelines and suggestions for parties that are involved in the CDS market in order to reduce the risk of trading the CDS and to reduce the asymmetric behaviour. The guidelines and suggestions concern the following involved parties:
a. ISDA
b. Corporate Borrowers c. CDS Investors
These sub-questions should guide the author to a satisfying answer on the main research question and subsequently help the author to accomplish the objective.
2.4 Conceptual
Model
In order to provide clarity and structure in the research a conceptual model has been developed. In line with De Leeuw’s “Bedrijskundige Methodologie” it is a representation of the author’s overall perception of the issues concerning this research. The lines and objects coloured in blue are the issues which are particularly addressed.
Exhibit 6 - Conceptual Model
The model is to represent not only the present, but also the anticipated future so that expected spreads can be derived and trades based on expectations can be put on.
In the model an arrow represents a causal relationship between two objects. For example the Corporate Borrower or Reference Entity (the company a CDS is related to) might create a corporate event (an acquisition, merger, refinancing etc.). Depending on the structure of this corporate event and subject to ISDA succession language a succession
event does or does not occur. That’s why succession event has a question mark written
Credit Risk Level
Determinants
Credit Default Swap
Spreads
Corporate Borrower
(Reference Entity)
Characteristics
ISDA Succession
Language
Capital Structure
Succession
Event?
Corporate
Event
Trades
Profit
behind it; there is usually not much clarity about the structure of an event and therefore uncertainty about a succession event occurring or not. If a succession event occurs the reference entity might be changed.
A straight line means that the lower object (in sense of the figure) is an attribute of the above object e.g. a corporate borrower has certain characteristics one of which is the
capital structure. This capital structure characteristic is especially pointed out because of
its importance for the credit risk level of a firm.
Similarly, the credit risk level is an attribute of the characteristics. The dashed circle is to illustrate that the credit risk level of a company is a very abstract variable. It depends on one’s perception and can be interpreted differently. These different perceptions enhance trading in products derived from credit risk, such as the CDS.
In the CDS theory (Hull and White (2000) and (2001)) the probability of default and the recovery rate determine CDS spreads. It is said that these variables represent the credit risk level. The two variables are influenced by various other factors in turn. Therefore the variables that are to represent the credit risk level of a given firm are simply labelled
determinants for CDS spreads. These determinants might also have a connection with
succession events. A dashed line is drawn between succession event? and determinants. A dashed line represents the research of this report. If the objective is satisfied one should be able to fill in the connection on the dashed line if there is any.
The other dashed line is between characteristics of a corporate borrower and succession
event? as this research will also address the possible relationship between the
characteristics of a corporate borrower and the likeliness of a succession event happening. Once the relationship between succession events and CDS spreads is determined, spreads should become more predictable and can be adjusted for the relationship. From these adjusted spreads, ING should be able to put on trades with less risk and hence more steady profits.
2.5
Limitations of the Research
The limitations of this research illustrate the boundaries of the scope of the research as well as other restrictions:
• Sovereign reference entities will be beyond the scope of this research as succession language on these entities differs from corporate entities.17 In addition, the market for sovereign credit default swaps is not as liquid as the market for corporate credit default swaps and therefore not as efficient.
• Legal issues will have an impact on this research, but as this research is initiated from a business perspective, it will not elaborate on the legal aspects. As such, the report may comment or make suggestions on ISDA 2003 language and
amendments under consideration. However, it will not focus on formulating new or additional language to the standard CDS contract.
• There might be other factors causing asymmetric behaviour besides succession language. This possibility will not be addressed in this research as this research focuses on issues involving (possible) succession events.
• As the CDS market is very new and succession issues concerning the CDS have only recently developed, data available on this issue is limited. Therefore this research will have a qualitative focus instead of a quantitative approach.
• This research is applicable on credit default swaps both in Europe and North America. Although this research might prove valuable for the Asian CDS, this is not the focus of the research.
• It is not the core purpose of this research to inform the reader on credit default swaps in general, therefore only the details of the CDS concerning succession issues are elaborated on. Other characteristics not relevant to this research may be addressed in a simplified manner or not mentioned at all.
• This research will be completed within six months.
2.6
Nature of the Research
Research can be structured in various ways. Baarda en de Goede (1997) distinguish four different types of research.
- Descriptive Research: Describes the appearance of one or more features within a specific group.
- Explorative Research: Discovering associations or differences among features within a specific group.
- Testing Research: Verifying an explanation or statement. This research is usually based on a hypothesis.
- Evaluative Research: A distinctive form of testing research. In this type a researcher has clear expectations of the results.
The nature and novelty of the problem of this thesis, clearly requires descriptive or explorative research; theories on succession issues haven’t yet been developed and it is hard to get a good overview of the problem with today’s knowledge.
Therefore the first part of this thesis will have descriptive features. The different features of succession issues will be discussed and some insight into the structure of the problem will be provided.
The objective of this thesis requires identification of associations or connections among the various concepts of the problem as well. Therefore, the second part of this thesis will have an explorative nature. It aims at developing a theory or formulation of hypotheses. Based on such theory or hypotheses a testing type of research can then be executed which shall
have a more quantitative nature. Exhibit 7 presents a graphical reproduction of the structure of this research.
Exhibit 7 - Research structure
The descriptive side is applicable on addressing the succession issues and making these issues known, whereas the explorative side will find its way in determining the relationship between the succession issues and CDS spreads.
Because this research is not evaluative it must use a wide approach to the problem and take all possible factors into account. In order to do this, the research is based on the findings of several diverse case studies which illustrate the span of the problem. This can be categorised as desk research.
Chapter 3
CDS spread determinants
Sub-question IChapter 4
Case Studies
Sub-question IIChapter 6
Generalisation
Sub-question III (cont’d)
Chapter 7
Guidelines for parties
Sub-question IV D DEESSCCRRIIPPTTIIVVEERREESSEEAARRCCHH EEXXPPLLOORRAATTIIVVEERREESSEEAARRCCHHChapter 5
Analysis
Sub-question III3 CDS
Spread
Determinants
In this chapter, the first sub-question will be addressed; the known determinants of CDS spreads. These determinants influence the perceived credit risk of a company, which in turn is reflected by CDS spreads. However, in case of asymmetric behaviour, the CDS spreads do not reflect the credit risk, hence the spreads are showing asymmetric behaviour in comparison with the determinants. To identify these situations, one must obviously address the determinants first.
Most research focusing on credit risk is devoted to corporate bonds. However, along with the growth of the CDS market, the need for research on the CDS evolves and bit by bit articles are being written. The transition from research on bonds to the CDS was encouraged by Collin-Dufresne, Goldstein and Martin (2001). They addressed determinants of credit spread changes instead of yield changes; the common dependent variable in prior research. Although they still focused on bonds, paragraph 1.3.4 shows that in general these credit spreads should equal CDS spreads, hence this research can be seen as a bridge to research on the CDS. Instead of determinants they write about proxies for both changes in the probability of future default and for changes in the recovery rate. In other words, the variables for calculating the price of a CDS (see section 1.3.3) are in practice represented by proxies (determinants). In their research they are not able to identify these proxies that significantly influence credit spreads. They found that the dominant component of credit spread changes was driven by local supply and demand shocks. However, they do identify some theoretical determinants of credit spread changes derived from structural models of default based on the work of Black and Scholes (1973) and introduced by Merton (1974). They mention that these models assume that an event of default is triggered when the firm’s value falls below some threshold, where the threshold is a function of the amount of outstanding debt. This means that according to these models, being the owner of a bond is correspondent to owning a similar risk-free bond with the sale of a put option on the issuing firm’s value (exercise price equal to the value of the risk-free bond). Using this knowledge as a basis, they identify six different theoretical determinants: Changes in the spot rate (risk free interest rate), changes in the slope of the yield curve, changes in leverage, changes in volatility, changes in the probability of a downward jump in firm value and changes in the business climate.
The first to explicitly address determinants of the CDS were Cossin & Hricko (2001): “There has not been any advanced study we are aware of that bear on Credit Risk as reflected in Credit Derivatives.” They also perform an empirical study in which they consider the following variables: Credit Ratings, Interest rate, Slope of the yield curve, Time-to-maturity, Stock prices, Variance or volatility of the firm’s assets, Leverage, Index returns and Idiosyncratic factors. Although they address more aspects by using more variables then Collin-Dufresne et al., the basis of the variables is more or less the same (some of them are even identical). Surprisingly enough compared to Collin-Dufresne et al.,
