The Analysis of the Influence of a Sovereign Foreign LT Issuer Credit Rating on Credit Spreads and Asset Swap Spreads

Contents

Introduction

1 RATINGS
1.1 Issuer Credit Rating Definitions
1.2 Issuer Credit Rating Criteria
1.3 Issuer Credit Rating Process
1.4 Issuer Credit Rating Examples
1.4.1 Moody’s
1.4.2 Standard & Poor’s

2 EFFICIENT MARKET HYPOTHESIS (EMH)

3 CORRELATION

4 GOVERNMENT YIELD BONDS AND SPREADS
4.1 THEORY
4.1.1 Yield Curve
4.1.2 Calculation Basis - Thomson Reuters Redemption Yields
4.2 CALCULATIONS PART I: Correlation between Credit Spreads and LT Credit Ratings
4.2.1 Calculation Methodology
4.2.2 Correlation Results
4.2.2.1 Portugal
4.2.2.2 Spain
4.2.2.3 Greece
4.2.3 Correlation Interpretation
4.3 CALCULATIONS PART II: Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating
4.3.1 Calculation and Methodology
4.3.2 Portugal Cases
4.3.2.1 Time period between 60 days before Rating Change to 20 days before Rating Change
4.3.2.2 Time period of 20 days before Rating Change
4.3.2.3 Time period of 20 days before Rating Change and 20 days after Rating Change
4.3.2.4 Time period of 20 days after Rating Change
4.3.2.5 Time period between 20 days after Rating Change to 60 days after Rating Change
4.3.3 Spain Cases
4.3.3.1 Time period between 60 days before Rating change to 20 days before Rating Change
4.3.3.2 Time period of 20 days before Rating Change
4.3.3.3 Time period of 20 days before Rating Change and 20 days after Rating Change
4.3.3.4 Time period of 20 days after Rating change
4.3.3.5 Time period between 20 days after Rating Change to 60 days after Rating Change
4.3.4 Greece Cases
4.3.4.1 Time period between 60 days before Rating Change to 20 days before Rating Change
4.3.4.2 Time period of 20 days before Rating Change
4.3.4.3 Time period of 20 days before Rating Change and 20 days after Rating Change
4.3.4.4 Time period of 20 days after Rating Change
4.3.4.5 Time period between 20 days after Rating Change to 60 days after Rating Change
4.4 CONCLUSION

5 ASSET SWAP SPREAD AND APPROXIMATION TO ASW
5.1 THEORY
5.1.1 Asset Swap Spread
5.1.2 Calculation Basis - Bloomberg Generic Bond Yields
5.2 CALCULATION PART I: Correlation between ASW and LT Issuer Credit Ratings
5.2.1 Correlation Methodology and Results
5.2.1.1 Portugal
5.2.1.2 Spain
5.2.1.3 Greece
5.3 CALCULATION PART II: Correlation between daily changes in ASW and changes in Foreign LT Credit Rating
5.3.1 Calculation and Methodology
5.3.2 Portugal Cases
5.3.2.1 Time period between 60 days before Rating Change to 20 days before Rating Change
5.3.2.2 Time period of 20 days before Rating Change
5.3.2.3 Time period of 20 days before Rating Change and 20 days after Rating Change
5.3.2.4 Time period of 20 days after Rating Change
5.3.2.5 Time period between 20 days after Rating Change to 60 days after Rating Change
5.3.3 Spain Cases
5.3.3.1 Time period between 60 days before Rating Change to 20 days before Rating Change
5.3.3.2 Time period of 20 days before Rating Change
5.3.3.3 Time period of 20 days before Rating Change and 20 days after Rating Change
5.3.3.4 Time period of 20 days after Rating Change
5.3.3.5 Time period between 20 days after Rating Change to 60 days after Rating Change
5.3.4 Greece Cases
5.3.4.1 Time period between 60 days before Rating Change to 20 days before Rating Change
5.3.4.2 Time period of 20 days before Rating Change
5.3.4.3 Time period of 20 days before Rating Change and 20 days after Rating Change
5.3.4.4 Time period of 20 days after Rating Change
5.3.4.5 Time period between 20 days after Rating Change to 60 days after Rating Change
5.3.5 CONCLUSION
5.4 CALCULATIONS PART III: Rolling correlations between daily changes in ASW and changes in LT Issuer Credit Ratings
5.4.1 Calculation Methodology
5.4.2 Portugal Cases
5.4.2.1 Rolling Correlation over 90 days, considering a time period of 20 days before Rating Change
5.4.2.2 Rolling correlation over 90 days, considering a time period of 20 days after Rating Change
5.4.3 Greece Cases
5.4.3.1 Rolling Correlation over 90 days, considering a time period of 20 days before Rating Change
5.4.3.2 Rolling Correlation over 90 days, considering a time period of 20 days after Rating Change

6 CONCLUSION

List of Figures

Figure 1 Municipal bond rating cheat sheets

Figure 2 Moody’s rating change for the Government of Portugal

Figure 3 Moody’s rating change for the Government of Spain

Figure 4 Moody’s rating change at the Government of Greece

Figure 5 S&P’s rating change for the Government of Portugal

Figure 6 S&P’s detailed rating change for the Government of Portugal

Figure 7 S&P’s rating change for the Government of Spain

Figure 8 S&P’s rating change for the Government of Greece

Figure 9 Yield curve shapes

Figure 10 Average redemption yield - datatype (RY)

Figure 11 Portugal Credit Spread & Foreign LT Issuer Credit Rating

Figure 12 Portugal Credit Spread & Foreign LT Issuer Credit Rating Correlation from 2002-2011

Figure 13 Spain Credit Spread & Foreign LT Issuer Credit Rating

Figure 14 Spain Credit Spread & Foreign LT Issuer Credit Rating Correlation from 2002-2011

Figure 15 Greece Credit Spread & Foreign LT Issuer Credit Rating

Figure 16 Greece Credit Spread & Foreign LT Issuer Credit Rating Correlation from 2002-2011

Figure 17 PT Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 60 days before to 20 days before Rating Change Event

Figure 18 PT Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before Rating Change Event

Figure 19 PT Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before to 20 days after Rating Change Event

Figure 20 PT Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after Rating Change Event

Figure 21 PT Correlation between daily changes in Credit Spread and changes in Foreign LT Credit

Rating: Analyzed time period of 20 days after to 60 days after Rating Change Event

Figure 22 ES Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 60 days before to 20 days before Rating Change Event

Figure 23 ES Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before Rating Change Event

Figure 24 ES Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before to 20 days after Rating Change Event

Figure 25 ES Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period 20 days after Rating Change Event

Figure 26 ES Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after to 60 days after Rating Change Event

Figure 27 GR Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 60 days before to 20 days before Rating Change Event

Figure 28 GR Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before Rating Change Event

Figure 29 GR Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before to 20 days after Rating Change Event

Figure 30 GR Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after Rating Change Event

Figure 31 GR Correlation between daily changes in Credit Spread and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after to 60 days after Rating Change Event

Figure 32 PT & ES - ASW vs. Yield-Swap rate (5Y and 10Y)

Figure 33 Correlations PT & ES - ASW vs Yield-Swap rate (5Y and 10Y)

Figure 34 Portugal Approximated ASW & Foreign LT Issuer Credit Rating

Figure 35 Portugal Approximated ASW & Foreign LT Issuer Credit Rating Correlation for 2002-2011

Figure 36 Spain Approximated ASW & Foreign LT Issuer Credit Rating

Figure 37 Spain Approximated ASW & Foreign LT Issuer Credit Rating Correlation for 2002-2011

Figure 38 Greece Approximated ASW & Foreign LT Issuer Credit Rating

Figure 39 Greece Approximated ASW & Foreign LT Issuer Credit Rating Correlation for 2002-2011

Figure 40 PT Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 60 days before to 20 days before Rating Change Event

Figure 41 PT Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before Rating Change Event

Figure 42 PT Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before to 20 days after Rating Change Event

Figure 43 PT Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after Rating Change Event

Figure 44 PT Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after to 60 days after Rating Change Event

Figure 45 ES Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 60 days before to 20 days before Rating Change Event

Figure 46 ES Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before Rating Change Event

Figure 47 ES Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before to 20 days after Rating Change Event

Figure 48 ES Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after Rating Change Event

Figure 49 ES Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after to 60 days after Rating Change Event

Figure 50 GR Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 60 days before to 20 days before Rating Change Event

Figure 51 GR Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before Rating Change Event

Figure 52 GR Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before to 20 days after Rating Change Event

Figure 53 GR Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after Rating Change Event

Figure 54 GR Correlation between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after to 60 days after Rating Change Event

Figure 55 PT Rolling Correlation over 90 days between daily changes in approximated ASW and changes in S&P’s Foreign LT Credit Rating: Analyzed time period of 20 days before Rating Change Event

Figure 56 PT Rolling Correlation over 90 days between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after Rating Change Event

Figure 57 GR Rolling Correlation over 90 days between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days before Rating Change Event

Figure 58 GR Rolling Correlation over 90 days between daily changes in approximated ASW and changes in Foreign LT Credit Rating: Analyzed time period of 20 days after Rating Change Event

AUSZUG

Ziel dieser Arbeit ist die Korrelation zwischen dem sogenannten „Foreign Long-Term Issuer Credit Rating“ (langfristigen Emittentenrating) und sowohl dem Kreditspread als auch dem Asset Swap Spread zu erfassen und hierbei die Rolle der Ratingagentur am Kapitalmarkt zu bemessen.

Ratings widerspiegeln die von der externen Ratingagentur eingeschätzte Bonität (Finanzstärke) und die Kreditwürdigkeit des Emittenten. Spreads bezeichnen die Erwartungen des Markets in Zusammenhang zum Risiko einer Investition. Demzufolge entschädigen hohe Spreads einen Investor für das eingegangene Risiko. Das Ergebnis der Analyse besagt, dass es tatsächlich einen Zusammenhang zwischen den langfristigen Emittentenrating und sowohl dem Credit Spread als auch dem Asset Swap Spread gibt. Ein hoher Spread korreliert stark mit einem niedrigen Kreditrating.

Obgleich der Zusammenhang besteht, so ist es dennoch nicht möglich ein klares Verhaltensmuster des Marktes zu identifizieren. In manchen, der analysierten Fälle gab es Marktbewegungen vor einer Ratingveränderung, was darauf hindeutet dass sowohl der Markt, als auch die Ratingagentur die gleichen Informationen berücksichtigten. In anderen Fällen wurde der Markt durch die Ratingveränderung beeinflusst. Und manchmal reagierte der Markt gegensätzlich zur Entscheidung der Ratingagentur und deren Erwartungen.

Allgemein reagiert der Markt unterschiedlich schnell und nicht homogen auf die „fundamentalen“ Informationen die er besitzt. Das macht es unmöglich klar auszusagen zu welchem Ausmaß Ratings einen Einfluss auf die Spreads ausüben. Das Resultat ist nicht klar genug um eine präzisere Antwort zu geben bzw. um das Marktverhalten gänzlich verstehen zu können. Der Markt ist ein hoch volatiles Umfeld in dem es unmöglich ist klar vorhersehbare Verhaltensmuster zu ziehen, welche sich lediglich auf der Kreditrating beziehen.

Die Unbeständigkeit der Reaktion ist teilweise widersprüchlich zur Theorie der starken Effizienzmarkthypothese, da der Markt ungleich auf den verfügbaren

Informationsstand reagiert bzw. andere Faktoren, welche in dieser Arbeit nicht behandelt werden, mit einpreist.

Abstract

The purpose of this paper is to measure the correlation between the foreign longterm issuer credit ratings and both, credit spreads and ASW and to assess the role that the Rating Agencies play on the capital markets.

Ratings reflect the financial strength and credit-worthiness of the issuer as assessed by the external rating agency. Spreads indicate the market’s expectations in connection to the riskiness of the investment. Hence high spreads compensate investors for the higher risk taken. The result of the analysis is that there indeed exist correlations between foreign long-term issuer credit ratings and both, credit spreads and ASW. A high spread is strongly correlated to a low credit rating.

However, even if the relationship exists, it is not possible to draw clear patterns of the market behavior. In some of the analyzed cases, market movements took place before a rating change, which indicates that both, the market and rating agencies considered the same information. In other cases, the market was influenced by the rating change. And sometimes the market reacted different to the rating agencies decisions and expectations.

Overall the market reacts differently fast and not homogenous to the “fundamental” information it has. This makes it impossible to clearly state the extent to which ratings influence spreads. The results are not clear enough to indicate a more precise answer or to fully understand the market behavior. The market is a highly volatile environment in which it is impossible to draw predictable behavior patterns in relation purely to the credit rating.

This inconsistency in reaction is partially conflicting the theory of the strong-form EMH, as the market reacts to the asymmetric level of information available or prices in other factors not covered in this paper.

Introduction

International capital markets environment is a very complex topic as there are multiple factors that influence a sovereign’s economic stability and consequently the yields on its bonds. These factors can be e.g. inflation, level of indebtedness, political stability etc. The aim of this dissertation is to analyze a special subarea of this complex system by taking out two specific variables to better understand the relationship between these two components and how they depend on each other. As a matter of course, this thesis is not explaining a complex interdependency by just one simple correlation between two variables. The analysis will answer the question if there is a correlation between credit spreads and asset swap spreads and foreign LT issuer credit ratings. The correlation is expected to be negative, having a coefficient close to -1.

If these two variables correlate it is important to understand how the interdependency exists and if its causality is given. Is the rating influencing the spread? Or to put it differently, is the Rating reflecting the situation of the market at a specific point or even influencing the market into a certain direction? Rating agencies were even accused to have partially provoked the crisis in 2008.

The answer to this question would clarify if the market is anticipating economic changes and consequently includes these into the price. Or if the market is being influenced by rating changes and reacts accordingly by an increase in price to compensate the investor for the deteriorating economic stability. The price always reflects the expectations of the market. Consequently a high yield on a bond indicates the risk level connected with the investment. On the contrary, a low price indicates a stable or lower risk connected with the investment.

In the course of this dissertation, it is analyzed if the strong-form Efficient Market Hypothesis is proved to be true. Following the EMH, the rating change should not influence the market to increase spreads. The market should already anticipate these changes, as all participants should share the same information on the market available. Moreover, a single market participant should not have the power to influence the market into a specific direction.

If the market reacted to rating changes it would imply that the information is asymmetrically spread and some market participant have indeed access to private information.

1 RATINGS

Internationally there are three rating agencies widely spread that issue ratings. A credit rating reflects the general creditworthiness of an obligor, or the creditworthiness of an obligor with respect to a particular debt security or other financial obligation as assessed by the external rating agency. Over the years credit ratings have achieved wide investor acceptance as convenient tools for differentiating credit quality.¹

1.1 Issuer Credit Rating Definitions

Issue ratings are an assessment of default risk, but may incorporate an assessment of relative seniority or ultimate recovery in the event of default.

A Standard & Poor's and Moody’s issuer credit rating is a forward-looking opinion about an obligor's overall financial capacity (its creditworthiness) to pay its financial obligations with an original maturity of one year or more. This opinion focuses on the obligor's capacity and willingness to meet its financial commitments as they come due. It does not apply to any specific financial obligation, as it does not take into account the nature of and provisions of the obligation, its standing in bankruptcy or liquidation, statutory preferences, or the legality and enforceability of the obligation. In addition, it does not take into account the creditworthiness of the guarantors, insurers, or other forms of credit enhancement on the obligation.² Such ratings reflect both the likelihood of default and any financial loss suffered in the event of default.³

Counterparty credit ratings, ratings assigned under the Corporate Credit Rating Service (formerly called the Credit Assessment Service) and sovereign credit ratings are all forms of issuer credit ratings. Issuer credit ratings can be either long term or short term. Long-term issuer credit ratings reflect the obligor's creditworthiness to honor financial obligations and contracts over a long-term time horizon.

illustration not visible in this excerpt

Figure 1 Municipal bond rating cheat sheets⁴

1.2 Issuer Credit Rating Criteria

Credit ratings agencies aim to play an important role in providing investors with an independent opinion about the creditworthiness of individual sovereigns. Ratings agencies want to help reduce the information asymmetry between issuers and investors. They believe ratings contribute to market efficiency, with ratings themselves available free of charge for investors, other market participants, and the public generally.⁵

Its sovereign rating criteria address the factors that in the agency’s opinion affect a sovereign government's willingness and ability to service its debt on time and in full. Our analysis focuses on a sovereign's performance over past economic and political cycles, as well as on factors that suggest to us greater or lesser fiscal and monetary flexibility over the course of future economic cycles.⁶

The five factors that form the foundation of our sovereign credit analysis are:

- Institutional effectiveness and political risks, reflected in the political score.
- Economic structure and growth prospects, reflected in the economic score.
- External liquidity and international investment position, reflected in the external score.
- Fiscal performance and flexibility, as well as debt burden, reflected in the fiscal score.
- Monetary flexibility, reflected in the monetary score.

1.3 Issuer Credit Rating Process

The rating process consists of several separate steps and typically includes a series of ongoing information exchanges between the ratings agency and the issuer. These interactions enable the agency to gather the information it needs to conduct its evaluation and form its rating opinion.⁷

For the rating process itself, the rating agency assigns an analytical team and appoints a committee. The analytical team examines the issuer's publicly reported financial information and any other relevant information provided by the issuer. At a so-called Management Meeting the analytical team meets with representatives of the issuer.⁸ In the case of a sovereign, they typically meet with representatives from the ministry of finance or the debt management office and other relevant departments like labour or social security. The analytical team also typically meets with central bank and private sector representatives. The purpose of the meeting is to enable the agency’s analysts to probe pertinent information in greater detail, including public information as well as other information that may be provided by the issuer.⁹

The final rating assigned by the committee is primarily determined by applying the rating criteria to the information that the analysts have collected and evaluated. The ratings consider the perceptions and insights of the analysts based on their consideration of all of the information they have obtained. The agency generally notifies the issuer of the rating and outlook, and provides reasons for the major factors supporting the rating. For public ratings, the agency publishes a press release announcing the final rating along with the reasons, distributes it to the media, and posts it on the internet.¹⁰

1.4 Issuer Credit Rating Examples

1.4.1 Moody’s

The tables below show Moody’s long term rating changes for Portugal, Spain and Greece over the last 25 years.

Moody's

Portugal, Government of

Rating Class History: LT Issuer Rating (Foreign)

26 July 2011 08:34:08 AM Eastern Standard Time

Date,Currency,Rating,Rating Action

05 Jul 2011,foreign,Ba2,Downgrade

05 Apr 2011,foreign,Baa1,Downgrade

05 Apr 2011,foreign,ON WATCH,Possible Downgrade

15 Mar 2011,foreign,A3,Downgrade

21 Dec 2010,foreign,ON WATCH,Possible Downgrade

13 Jul 2010,foreign,A1,Downgrade

05 May 2010,foreign,ON WATCH,Possible Downgrade

04 May 1998,foreign,Aa2,Upgrade

10 Feb 1997,foreign,Aa3,Upgrade

18 Nov 1986,foreign,A1,New

© 2011 Moody's Investors Service, Inc. and/or its licensors and affiliates (collectively, "MOODY'S"). All rights reserved.

Figure 2 Moody’s rating change for the Government of Portugal¹¹

Moody's

Spain, Government of

Rating Class History: LT Issuer Rating (Foreign)

26 July 2011 08:34:46 AM Eastern Standard Time

Date,Currency,Rating,Rating Action

10 Mar 2011,foreign,Aa2,Downgrade

15 Dec 2010,foreign,ON WATCH,Possible Downgrade

30 Sep 2010,foreign,Aa1,Downgrade

30 Jun 2010,foreign,ON WATCH,Possible Downgrade

13 Dec 2001,foreign,Aaa,Upgrade

19 Sep 2001,foreign,ON WATCH,Possible Upgrade

03 Feb 1988,foreign,Aa2,New

© 2011 Moody's Investors Service, Inc. and/or its licensors and affiliates (collectively,

"MOODY'S"). All rights reserved.

Figure 3 Moody’s rating change for the Government of Spain¹²

Moody's

Greece, Government of

Rating Class History: LT Issuer Rating (Foreign)

26 July 2011 08:28:40 AM Eastern Standard Time

Date,Currency,Rating,Rating Action

25 Jul 2011,foreign,Ca,Downgrade

01 Jun 2011,foreign,Caa1,Downgrade

09 May 2011,foreign,ON WATCH,Possible Downgrade

07 Mar 2011,foreign,B1,Downgrade

16 Dec 2010,foreign,ON WATCH,Possible Downgrade

14 Jun 2010,foreign,Ba1,Downgrade

22 Apr 2010,foreign,ON WATCH,Possible Downgrade

22 Apr 2010,foreign,A3,Downgrade

22 Dec 2009,foreign,A2,Downgrade

29 Oct 2009,foreign,ON WATCH,Possible Downgrade

04 Nov 2002,foreign,A1,Upgrade

14 Jul 1999,foreign,A2,Upgrade

20 Feb 1998,foreign,ON WATCH,Possible Downgrade

23 Dec 1996,foreign,Baa1,Upgrade

04 Nov 1996,foreign,ON WATCH,Possible Upgrade

24 May 1994,foreign,Baa3,Downgrade

19 Jul 1990,foreign,Baa1,New

© 2011 Moody's Investors Service, Inc. and/or its licensors and affiliates (collectively, "MOODY'S"). All rights reserved.

Figure 4 Moody’s rating change at the Government of Greece¹³

1.4.2 Standard & Poor’s

The graphs below show Standard & Poor’s long term rating changes for Portugal, Spain and Greece over the last 25 years.

illustration not visible in this excerpt

Figure 5 S&P’s rating change for the Government of Portugal¹⁴

illustration not visible in this excerpt

Figure 6 S&P’s detailed rating change for the Government of Portugal¹⁵

illustration not visible in this excerpt

Figure 7 S&P’s rating change for the Government of Spain¹⁶

illustration not visible in this excerpt

Figure 8 S&P’s rating change for the Government of Greece¹⁷

2 EFFICIENT MARKET HYPOTHESIS (EMH)

The Efficient Market Hypothesis (EMH) is based on the assumption that prices of securities in financial markets fully reflect all available information, public and private.¹⁸ This implies that no group of investors (or market participants) has access to private information and no market participant has monopolistic access to information relevant to the formation of prices. The strong-form EMH extends the assumption of efficient markets, in which prices adjust rapidly to the release of new public information, to assume perfect markets, in which all information is cost-free and available to everyone at the same time.¹⁹

Following the strong-from EMH, a rating change should not influence the market to increase spreads. The market should already anticipate rating changes, as all participants should share the same information on the market available. Moreover, a single market participant should not have the power to influence the market into a specific direction.

In course of this dissertation, it is analyzed if the market behavior is in contradiction to this hypothesis. If the market reacted to changes, it would imply that the information is asymmetrically spread.

3 CORRELATION

The correlation is one of the most familiar measures in statistics. It is a number that describes the degree of relationship between two variables.²⁰ The most common correlation technique is called the Pearson or product-moment correlation. Like all statistical techniques, correlation is only appropriate for certain kinds of data. Correlation works for quantifiable data in which numbers are meaningful, usually quantities of some sort. It cannot be used for purely categorical data, such as gender, brands purchased, or favorite color.²¹

When measured in a population the Pearson Product Moment correlation is designated by the Greek letter rho (ρ). When computed in a sample, it is designated by the letter "r" and is sometimes called "Pearson's r." Pearson's correlation reflects the degree of linear relationship between two variables. It ranges from +1 to -1. A correlation of +1 stands for a perfect positive linear relationship between the variables. A correlation of -1 stands for a perfect negative linear relationship (anticorrelation) between the variables. The closer the coefficient is to either −1 or 1, the stronger the correlation between the variables. As it approaches zero there is less of a relationship (closer to uncorrelated).²² A non-zero correlation means that the numbers are related, but unless the coefficient is either 1 or -1 there are other influences and the relationship between the two numbers is not fixed. So if you know one number you can estimate the other, but not with certainty. The closer the correlation coefficient is to zero the greater the uncertainty. A correlation of 0 means there is no linear relationship between the two variables.²³

Correlations are rarely if ever 0, 1, or -1. Often they only indicate whether correlations are negative or positive.²⁴

The correlation coefficient of data set can be derived by dividing the covariance of the two variables by the product of their standard deviations. The population correlation coefficient ρ X,Y between two random variables X and Y with expected values μ X and μ Y and standard deviations σ X and σ Y is defined as:

illustration not visible in this excerpt

R is the expected value operator, cov means covariance, and, corr a widely used alternative notation for Pearson's correlation. ²⁶

4 GOVERNMENT YIELD BONDS AND SPREADS

4.1 THEORY

4.1.1 Yield Curve

The yield is the annual rate of return on an investment, expressed as a percentage.²⁷ The yield to maturity is the percentage rate of return paid if the security is held toits maturity date. The calculation is based on the coupon rate, length of time to maturity, and market price. It assumes that coupon interest paid over the life of the security is reinvested at the same rate.²⁸

The yield curve is a line that plots the interest rates, at a set point in time, of bonds having equal credit quality, but differing maturity dates, ranging from shortest to longest.²⁹ It is also called "term structure of interest rates".

The yield curve enables investors at a quick glance to compare the yields offered by short-term, medium-term and long-term bonds. Note that it does not plot coupon rates against a range of maturities -- that's called a spot curve³⁰

The yield curve can take three primary shapes. If short-term yields are lower than long-term yields (the line is sloping upwards), then the curve is referred to a positive (or "normal") yield curve. If short-term yields are higher than long-term yields (the line is sloping downwards), then the curve is referred to as an inverted (or "negative") yield curve. Finally, a flat yield curve exists when there is little or no difference between short- and long-term yields.³¹

illustration not visible in this excerpt

Figure 9 Yield curve shapes³²

4.1.2 Calculation Basis - Thomson Reuters Redemption Yields

For the relevant calculations in this thesis, the Redemption Yields (RY) or Yield to maturity (YTM) of Portugal, Spain, Greece and Germany are relevant. The RY data that is given for TRBD10T (TR GERMANY GVT BMK BID YLD 10Y (E)) represents the Native Yield values that was extract of Thomson Reuters’ information system.³³

The individual yield is not treated as a “generic” yield but a “native” yield which is calculated using the issuer defined day-count convention and the market specific compounding rule.

For the Thomson Reuters Benchmark, the criterion for selecting a benchmark is not only the year the bond matures in. The main factors are market participant consensus, as well as maturity and liquidity factors. The following selection describes the criteria used to select a benchmark:³⁴

After research and after consulting with local sources, Reuters’ analysts maintain lists of bonds that appear as the default benchmark securities for the major bond markets. A Reuters-selected benchmark security is the most liquid issue and has the highest turnover in its maturity range. It tends to be the most recent issue of good size, the terms of which set standards for the market.

Some other determining factors for selecting a benchmark:

- Can be an older issue.
- Pricing must be readily available.
- Must be clear of options or credit penalty.

In the process of selecting default benchmark securities, Reuters analysts:

- Canvas market markers for new benchmark changes.
- Obtain market consensus for our selections.
- Consult with central banks or treasuries.
- Monitor trading activities of selected benchmarks and other choices (trade volumes, yields, bid/ask spreads, closeness to par value and so on).
- Usually populate terms with only the market-accepted benchmarks. In cases where there are other liquid bonds that the market might use within the yield curve, we might use those as well, especially for countries with only one or two market benchmarks. Hence, if newly issued bond falls under the above criteria’s, it will be used as a benchmark.

The values extracted from Thomson Reuters Benchmark are End of Day (EOD) yield data.

The average redemption yield RY of an index at time t is calculated by the following where the summations are over the bonds currently in the index:

illustration not visible in this excerpt

Figure 10 Average redemption yield - datatype (RY)³⁵

The calculations for the Datastream government bond indices follow the recommendations of the EFFAS Index sub-committee and use the following notations:

illustration not visible in this excerpt

A Accrued interest to the 'normal' settlement date

N Nominal value of amount outstanding if known, otherwise the issued amount Y Redemption yield to assumed maturity

L Life to assumed maturity

D Duration

X Convexity

C Coupon rate %

Value of any coupon payment received from the ith bond at time t or since time (t-1). If none, the value = 0

R Redemption price of the bond

P* For any non-serial bond this will be the same as P.

4.2 CALCULATIONS PART I:

Correlation between Credit Spreads and LT Credit Ratings

4.2.1 Calculation Methodology

4.2.2 Correlation Results

The first graphs below show the credit spreads in different time buckets, from 3M (3 months maturity) to 30Y (30 years maturity). Spreads are calculated as a subtraction of a government yield bond minus the Germany government yield bond in the same bucket of the same day, as Germany government bond is commonly referred to as the risk free rate.

Moreover these graphs show the foreign long-term issuer credit ratings of the agencies Standard & Poor’s and Moody’s, translated into a scale from 1 to 22. The higher the number, the better the rating is. The considered time period is January 2002 to August 2011.

The second graphs show the linear correlation between credit spreads and Moody’s or S&P’s foreign long term rating.

Short term credit spreads are expected to correlate less with the long term credit rating than longer term credit spreads, as the rating refers to a long term rating.

4.2.2.1 Portugal

illustration not visible in this excerpt

Figure 11 Portugal Credit Spread & Foreign LT Issuer Credit Rating³⁶

The graph visualizes the impacts of the crisis on Portugal. Spreads, i.e. the sovereign’s refinancing costs, rose up to 20,037% (3Y on 18.07.2011).

The graphs below show the calculated linear correlation between redemption yield spreads and Moody’s or S&P’s foreign long term rating.

illustration not visible in this excerpt

Figure 12 Portugal Credit Spread & Foreign LT Issuer Credit Rating Correlation from 2002-2011

[...]

---

¹ Standard & Poor’s (2012) online

² Standard & Poor’s (2012): RatingDirect on the Global Credit Portal pp 3

³ Moody’s (2012): Rating Sources and Definitions pp. 4

⁴ Municipalbond (2012) online

⁵ Standard & Poor’s (2012): How We Rate Sovereigns pp 3

⁶ ibid

⁷ ibid

⁸ Moody’s (2012): online

⁹ Standard & Poor’s (2012): How We Rate Sovereigns pp 6

¹⁰ ibid

¹¹ Moody’s (2012) online (restricted permission)

¹² ibid

¹³ ibid

¹⁴ S&P’s (2012) online (restricted permission)

¹⁵ S&P’s (2012) online (restricted permission)

¹⁶ ibid

¹⁷ ibid

¹⁸ Frederic S. Mishkin (2010): The Economics of Money, Banking and Financial Markets p. 156

¹⁹ Brown/Reilly (2009): Analysis of Investments and Management of Portfolios pp 153

²⁰ Socialresearchmethods (2012): online

²¹ Surveysystem (2012): online

²² Everything explained (2012): online

²³ Moneyterms (2012): online

²⁴ Everything explained (2012): online

²⁵ R is the expected value operator, cov means covariance, and, corr a widely used alternative notation for Pearson's correlation.²⁶

²⁵ Ncalculators (2012): online

²⁶ Mihaela Ioana Baritz and Diana Laura Cotoros (2011): Development of correlative investigation technique (CIT) of human performances in specific conditions of postural malfunctions p1

²⁷ Investorwords (2012) online

²⁸ Harald Schlick (2012): ALM Expert in Bank Austria

²⁹ Investopedia (2012) online

³⁰ Investorwords (2012): online

³¹ ibid

³² ibid

³³ Thomson Reuters Customer Support (2011)

³⁴ ibid

³⁵ Datastream (2012) online

³⁶ Figures: Reuters (2012), S&P’s (2012) and Moody’s (2012) online (restricted permission)