Implementation of a Performance Attribution System in a Funds Management Company
With Regard to Support Controlling and Marketing
©2004
Masterarbeit
153 Seiten
Zusammenfassung
Inhaltsangabe:Introduction:
Within this thesis fundamentals of performance attribution are treated as well as the practical implementation of such a system. However, for many people who already asked to receive a copy of this work, the most important part will be the analysis of the (partly internet based) questionnaire revolving around performance attribution.
The theoretical part contains the fundamentals of portfolio theory and the constructive CAPM, followed by basics of portfolio management. Next topics are the estimation of risk and performance for investment portfolios. The finishing touch is done by introduction of performance attribution theories handling both, equity and bond attribution.
Within the practical part a short analysis of the affected asset management company is performed, accompanied by research of performance presentation.
A questionnaire has been sent to 80 non-banks of the German HDAX index as well as to all members of the BVI. Furthermore an internet based version of the questionnaire has been introduced to a large number of institutional clients. The analysis of this questionnaire is one of the topics of this work.
For the affected company the results from the questionnaire are used to identify the competitors possibilities and the clients needs, resulting in fundamental ideas for the catalogue of requirements.
The analysis of the investment process and based on this the choice and the implementation of a performance attribution system are the last parts of this work. Still it has to be mentioned that recommendations for organizational changes are part of implementation.
There are three reasons why this thesis has been prepared.
The first one is quite trivial. As performance attribution is more and more present in the market, deeper knowledge of it shall be gained for the affected company.
Primarily the department for quantitative research and information technology has to come up with this knowledge to act as a multiplier and to ascertain technological and / or organizational requirements for its implementation.
Another scope of this work is to determine the actual status of performance attribution performed by competitors. To get this information competition is referred to and likewise clients of competitors are interviewed.
To benefit from this thesis not only the theoretical background will be looked upon, but there will be approaches to completely integrate performance attribution.
Proposed usage will […]
Within this thesis fundamentals of performance attribution are treated as well as the practical implementation of such a system. However, for many people who already asked to receive a copy of this work, the most important part will be the analysis of the (partly internet based) questionnaire revolving around performance attribution.
The theoretical part contains the fundamentals of portfolio theory and the constructive CAPM, followed by basics of portfolio management. Next topics are the estimation of risk and performance for investment portfolios. The finishing touch is done by introduction of performance attribution theories handling both, equity and bond attribution.
Within the practical part a short analysis of the affected asset management company is performed, accompanied by research of performance presentation.
A questionnaire has been sent to 80 non-banks of the German HDAX index as well as to all members of the BVI. Furthermore an internet based version of the questionnaire has been introduced to a large number of institutional clients. The analysis of this questionnaire is one of the topics of this work.
For the affected company the results from the questionnaire are used to identify the competitors possibilities and the clients needs, resulting in fundamental ideas for the catalogue of requirements.
The analysis of the investment process and based on this the choice and the implementation of a performance attribution system are the last parts of this work. Still it has to be mentioned that recommendations for organizational changes are part of implementation.
There are three reasons why this thesis has been prepared.
The first one is quite trivial. As performance attribution is more and more present in the market, deeper knowledge of it shall be gained for the affected company.
Primarily the department for quantitative research and information technology has to come up with this knowledge to act as a multiplier and to ascertain technological and / or organizational requirements for its implementation.
Another scope of this work is to determine the actual status of performance attribution performed by competitors. To get this information competition is referred to and likewise clients of competitors are interviewed.
To benefit from this thesis not only the theoretical background will be looked upon, but there will be approaches to completely integrate performance attribution.
Proposed usage will […]
Leseprobe
Inhaltsverzeichnis
ID 8669
Vogel, Marc: Implementation of a Performance Attribution System in a Funds
Management Company - In Regard to Support Controlling and Marketing
Hamburg: Diplomica GmbH, 2005
Zugl.: Hogeschool Zeeland, MA-Thesis / Master, 2004
Dieses Werk ist urheberrechtlich geschützt. Die dadurch begründeten Rechte,
insbesondere die der Übersetzung, des Nachdrucks, des Vortrags, der Entnahme von
Abbildungen und Tabellen, der Funksendung, der Mikroverfilmung oder der
Vervielfältigung auf anderen Wegen und der Speicherung in Datenverarbeitungsanlagen,
bleiben, auch bei nur auszugsweiser Verwertung, vorbehalten. Eine Vervielfältigung
dieses Werkes oder von Teilen dieses Werkes ist auch im Einzelfall nur in den Grenzen
der gesetzlichen Bestimmungen des Urheberrechtsgesetzes der Bundesrepublik
Deutschland in der jeweils geltenden Fassung zulässig. Sie ist grundsätzlich
vergütungspflichtig. Zuwiderhandlungen unterliegen den Strafbestimmungen des
Urheberrechtes.
Die Wiedergabe von Gebrauchsnamen, Handelsnamen, Warenbezeichnungen usw. in
diesem Werk berechtigt auch ohne besondere Kennzeichnung nicht zu der Annahme,
dass solche Namen im Sinne der Warenzeichen- und Markenschutz-Gesetzgebung als frei
zu betrachten wären und daher von jedermann benutzt werden dürften.
Die Informationen in diesem Werk wurden mit Sorgfalt erarbeitet. Dennoch können
Fehler nicht vollständig ausgeschlossen werden, und die Diplomarbeiten Agentur, die
Autoren oder Übersetzer übernehmen keine juristische Verantwortung oder irgendeine
Haftung für evtl. verbliebene fehlerhafte Angaben und deren Folgen.
Diplomica GmbH
http://www.diplom.de, Hamburg 2005
Printed in Germany
To those
who love me,
who supported my studies,
who suffered from my absence,
who preferred playing with me in the afternoon and
who entered my study room whenever they thought I am home
my wife Nicole,
my daughter Susan
and my sons Tim
and Jan
Implementation of a Performance Attribution System
in a Funds Management Company
In Regard to Support Controlling and Marketing
by Marc Vogel
20/7/0
4
73 75
77 79
81 83
85 87
89 91
93 95
97 99
01 03
-3
-2
-1
0
1
2
3
4
5
6
-10
-5
0
5
10
15
20
25
30
CAPE
X NO
N FAR
M %Y
/Y
PRIVA
TE FIX
ED IN
V ST %
Y /Y(R
ECHT
E SKA
LA)
Sourc
e: DA
TAST
REAM
20/7/0
4
73 75
77 79
81 83
85 87
89 91
93 95
97 99
01 03
-3
-2
-1
0
1
2
3
4
5
6
-10
-5
0
5
10
15
20
25
30
CAPE
X NO
N FAR
M %Y
/Y
PRIVA
TE FIX
ED IN
V ST %
Y /Y(R
ECHT
E SKA
LA)
Sourc
e: DA
TAST
REAM
i
I
List of Abbreviations
X
The Riemann sum more correctly written as
=
n
i
i
X
1
is being used in short form in this Master Thesis, where
confusion is unlikely.
AIMR
The Association for Investment Management and Research
BHB model
Performance attribution model according to Brinson, Hood and
Beebower
BVI
(German) Bundesverband Deutscher Investmentmanager
German committee of investment managers
CAPM
Capital Asset Pricing Model
DVFA
(German)
Deutsche Vereinigung für Finanzanalyse und Asset
Management
German Society for Financial Analysis and Asset Management
FERI
Feri Finance Group,
Counsellor for Investment Management
GIPS
Global Investment Performance Standards, initiated by AIMR
HDAX
(German) Deutscher Aktienindex Einhundert
German equity Index with top 100 equities
HSBC Trinkaus
HSBC Trinkaus & Burkhardt KGaA
Former German private bank, 70% owned by HSBC Group;
Landesbank Baden Württemberg has a 20% stake.
HTCM
HSBC Trinkaus Capital Management GmbH
Asset advisory company of HSBC Trinkaus.
INKA
(German)
Internationale Kapitalanlagegesellschaft
Investment Company of HSBC Trinkaus
ii
ISMA
International Securities Market Association
Formula for the calculation of bond price according to ISMA.
KAG
(German)
Kapitalanlagegesellschaft
Investment company according to German law.
NPV
Net present value
P/E
Price / earnings ratio
PAS
Performance Attribution System
PPS
Performance Presentation Standards
VaR
Value at Risk
iii
II List of Content
I
List of Abbreviations...i
II List of Content ...iii
III
List of Figures...viii
1
Introduction...1
1.1
Executive Summary...1
1.2
Scope of Work...2
2
Implementation of a Performance Attribution System ...3
2.1
Definition of Problem ...3
2.2
Relevance of Performance Attribution ...5
2.3
Research Methods ...6
2.3.1
Analysis of Literature ...6
2.3.2
Questionnaire ...7
2.3.3
Interviews ...7
2.3.4
Conclusion...8
3
Theory ...9
3.1
Analysis of Relevant Theories ...9
3.1.1
Basic Principles of Asset Management ...9
3.1.1.1 Portfolio Theory according to Markowitz...9
3.1.1.2 Capital Asset Pricing Model (CAPM)...10
3.1.1.3 Multi Beta CAPM...10
3.1.1.4 Arbitrage Pricing Theory ...10
3.1.1.5 Market Model...11
3.1.1.6 Information Efficiency ...11
3.1.2
Risk Measurement ...12
3.1.2.1 Volatility...12
3.1.2.2 Value at Risk...12
3.1.2.3 Probability of Failure...12
3.1.2.4 Asymmetric Risk Values...12
3.1.2.5 Duration...13
3.1.3
Performance Measurement...13
3.1.3.1 Performance as Return ...13
3.1.3.1.1 Return Estimation of Securities...14
iv
3.1.3.1.2 Money weighted Rate of Return...14
3.1.3.1.3 Time weighted Rate of Return...15
3.1.3.1.4 BVI Method ...15
3.1.3.1.5 Modified Dietz Method...15
3.1.3.1.6 Dietz Method ...15
3.1.3.2 Performance as Risk weighted Return ...15
3.1.3.2.1 Sharpe Ratio...16
3.1.3.2.2 Treynor Ratio...16
3.1.3.2.3 Jensen Ratio ...16
3.1.3.2.4 Tracking Error ...16
3.1.3.2.5 Differential Return ...17
3.1.3.2.6 Risk Adjusted Performance ...17
3.1.3.2.7 Information Ratio ...17
3.1.3.2.8 Sortino Ratio...17
3.1.4
Performance Contribution...17
3.1.5
Performance Attribution...18
3.1.5.1 Brinson Hood Beebower ...18
3.1.5.2 Brinson / Fachler...18
3.1.5.3 Karnosky Singer ...18
3.1.5.4 Geometric Approach ...19
3.1.5.5 Fixed Income Attribution...19
3.1.5.6 Linking across Time ...19
3.1.6
Fund Management ...19
3.1.6.1 Active / Passive Management...19
3.1.6.2 Other Management Styles ...20
3.2
Selection of Appropriate Theories ...20
3.2.1
Portfolio Selection according to Markowitz ...20
3.2.2
Capital Asset Pricing Model ...23
3.2.2.1 Capital Market Line...23
3.2.2.2 Equity Line...25
3.2.3
Basic Principles of Portfolio Management...28
3.2.3.1 The Investment Process...28
3.2.3.1.1 Investment Aim ...29
3.2.3.1.2 Specific Risk...29
v
3.2.3.1.3 Asset Liability...29
3.2.3.1.4 Policy Statement ...30
3.2.3.2 Portfolio Management ...31
3.2.3.2.1 Asset Allocation ...31
3.2.3.2.2 Passive Management ...31
3.2.3.2.3 Active Management ...33
3.2.4
Estimation of Risk ...34
3.2.4.1 Volatility...34
3.2.4.2 Duration...35
3.2.4.2.1 Macaulay Duration...36
3.2.4.2.2 Modified Duration...38
3.2.5
Estimation of Performance...38
3.2.5.1 Money weighted Rate of Return...39
3.2.5.2 Time weighted Rate of Return...40
3.2.5.3 Comparison of Time weighted and Money weighted Return
41
3.2.5.4 BVI Method ...42
3.2.5.5 Modified Dietz Method...43
3.2.6
Performance as Risk weighted Return...44
3.2.6.1 Sharpe Ratio...44
3.2.6.2 Treynor Ratio ...45
3.2.6.3 Jensen Ratio ...46
3.2.6.4 Tracking Error...47
3.2.7
Contribution ...48
3.2.8
Performance Attribution, Arithmetic...51
3.2.8.1 Definitions...51
3.2.8.2 The Brinson, Hood, Beebower Approach...52
3.2.8.3 Conclusions given by the BHB Approach ...53
3.2.8.4 The Brinson Fachler Approach...54
3.2.8.5 Linking across Time ...56
3.2.9
Performance Attribution, Geometric ...59
3.2.9.1 The Model ...60
3.2.9.2 Linking across Time ...62
vi
3.2.9.3 Comparison of Arithmetic and Geometric Performance Attribution
62
3.2.10
Fixed Income Attribution...63
3.2.10.1
Differences between Equity and Fixed Income Attribution
63
3.2.10.2
Bond Attribution according to Steve Campisi...64
3.2.10.3
Bond Attribution according to Tim Lords ...65
4
Case ...67
4.1
Relevance of Problem for HTCM / Case Description ...67
4.1.1
Analysis of HTCM...67
4.1.1.1 Organisational Structure...68
4.1.1.2 Vision and Mission of HTCM ...69
4.1.1.3 SWOT Analysis ...69
4.1.2
PPS...70
4.1.2.1 GIPS...71
4.1.2.2 DVFA PPS ...74
4.1.3
Motivation ...74
4.1.4
Competitive Background...75
4.2
How to Apply Research Methods to the Case...75
4.2.1
Practical Relevance of Theories...75
4.2.2
Relevance of PAS for Marketing / Controlling ...76
4.3
How to apply selected Methods ...76
4.3.1
Specifications of PAS according to Literature...76
4.3.2
Particular Specifications according to HTCM ...77
4.4
Analysis of Questionnaire...78
4.4.1
Detailed Explanation of Questionnaire...78
4.4.1.1 Estimation of Company Sizes...78
4.4.1.2 Evaluation of the actual State ...79
4.4.1.2.1 Present Estimation and Presentation of Return ...79
4.4.1.2.2 Detailed Analysis of current Asset Attribution ...80
4.4.1.3 Evaluation of the Client's Future Decisions ...84
4.4.2
Results of the Questionnaire ...85
4.4.2.1 Countries, Sectors and Sizes of participating Companies...
87
vii
4.4.2.2 Evaluation of the actual State of Performance Attribution..
89
4.4.3
Proposal for Specifications resulting from the Questionnaire .
105
4.5
How to solve the Problem...106
4.5.1
Investigation of Investment Process ...106
4.5.1.1 Asset Allocation...106
4.5.1.2 Equity Investment Process ...107
4.5.1.3 Fixed Income Investment Process ...109
4.5.2
Creating the Catalogue of Requirements...111
4.5.2.1 Common Aspects...112
4.5.2.1.1 The Software Vendor ...112
4.5.2.1.2 Software Technology ...112
4.5.2.1.3 Plausible Functionality...113
4.5.2.1.4 Customizability of the System...113
4.5.2.2 Specific Aspects...113
4.5.3
Screening of PAS Software (Overview)...114
4.5.4
Selected PAS...115
4.5.5
SMART Objectives and Parameters...119
4.5.6
Proposed Usage in Controlling ...120
4.5.7
Proposed Usage in Marketing ...122
5
Recommendation & Conclusion...124
5.1
Conclusion ...124
5.2
Recommendation...125
6
Literature...i
6.1
Books / Journals ...i
6.2
Internet ...v
7
Affirmation ...vii
8
Attachments ... A
viii
III List of Figures
Figure 1: Criteria for the choice of Investment funds...3
Figure 2: Value Chain for Asset Management...6
Figure 3: Structure of Research Methods...8
Figure 4: Efficiency curve for market portfolios ...22
Figure 5: Capital Market Line...24
Figure 6: The Portfolio Management Process...28
Figure 7: Scheme for the Brinson Approach...52
Figure 8: Calculation Instruction for the Brinson Approach ...52
Figure 9: Schema for the Brinson Fachler Approach...56
Figure 10: Steve Campisi's decomposition of Bond Return 1...64
Figure 11: Steve Campisi's decomposition of Bond Return 2...65
Figure 12: Tim Lord's decomposition of Bond Return...66
Figure 13: Organisational structure of HTCM...68
Figure 14: SWOT Analysis of HTCM ...70
Figure 15: Elements of GIPS ...72
Figure 16: Participants in Questionnaire...86
Figure 17: Country Analysis ...87
Figure 18: Sector Analysis ...87
Figure 19: KAG Assets under Management ...88
Figure 20: Amount of Assets ...88
Figure 21: Estimation of Return ...89
Figure 22: Mathematical Methods ...90
Figure 23: Investment Strategy...91
Figure 24: Asset Classes for Performance Attribution...92
Figure 25: Cash Benchmark...92
Figure 26: Equity Attribution Model...93
Figure 27: Fixed Income Model ...94
Figure 28: Existence of a Residual ...95
Figure 29: Source of Exchange Rates...96
Figure 30: Company Features of the Attribution Model...96
Figure 31: KAG Features of the Attribution Model ...97
Figure 32: Priority of Performance Attribution ...98
ix
Figure 33: Relevance of Subjects - Personal relationship...99
Figure 34: Relevance of Subjects Other Relation to Holding...100
Figure 35: Relevance of Subjects Previous Performance ...100
Figure 36: Relevance of Subjects Dimension of Asset Manager ...101
Figure 37: Relevance of Subjects Reputation of Asset Manager ...101
Figure 38: Relevance of Subjects Financial Concept...102
Figure 39: Relevance of Subjects Competency in Risk Analysis...102
Figure 40: Relevance of Subjects Performance Attribution System ...103
Figure 41: Relevance of Subjects - Comparison...104
Figure 42: Asset Classes Relevant for Performance Attribution...104
Figure 43: Quality of Performance Attribution...105
Figure 44: Investment Process Equities 1 ...107
Figure 45: Investment Process Equities 2 ...108
Figure 46: Investment Process Equities 3 ...108
Figure 47 Investment Process Bonds 1 ...109
Figure 48: Investment Process Bonds 2 ...110
Figure 49 Investment Process Bonds 3 ...110
Figure 50 Investment Process Bonds 4 ...111
Figure 51: Yield Curve Change ...116
Figure 52: Shift Movement of the Yield Curve ...117
Figure 53: Twist Movement of the Yield Curve...117
Figure 54: Butterfly Movement of the Yield Curve ...117
Figure 55: Stage wise Model of the Investment Process...121
Figure 56: Evolutionary Model of the Investment Process...121
1
1 Introduction
1.1 Executive Summary
Within this thesis fundamentals of performance attribution are treated as well as the
practical implementation of such a system. However, for many people who already
asked to receive a copy of this work, the most important part will be the analysis of the
(partly internet based) questionnaire revolving around performance attribution.
The theoretical part contains the fundamentals of portfolio theory and the constructive
CAPM, followed by basics of portfolio management. Next topics are the estimation of
risk and performance for investment portfolios. The finishing touch is done by
introduction of performance attribution theories handling both, equity and bond
attribution.
Within the practical part a short analysis of the affected asset management company is
performed, accompanied by research of performance presentation.
A questionnaire has been sent to 80 non-banks of the German HDAX index as well as
to all members of the BVI. Furthermore an internet based version of the questionnaire
has been introduced to a large number of institutional clients. The analysis of this
questionnaire is one of the topics of this work.
For the affected company the results from the questionnaire are used to identify the
competitors' possibilities and the clients' needs, resulting in fundamental ideas for the
catalogue of requirements.
The analysis of the investment process and based on this the choice and the
implementation of a performance attribution system are the last parts of this work. Still
it has to be mentioned that recommendations for organizational changes are part of
implementation.
2
1.2 Scope of Work
There are three reasons why this thesis has been prepared.
The first one is quite trivial. As performance attribution is more and more present in the
market, deeper knowledge of it shall be gained for the affected company.
Primarily the department for quantitative research and information technology has to
come up with this knowledge to act as a multiplier and to ascertain technological and /
or organizational requirements for its implementation.
Another scope of this work is to determine the actual status of performance attribution
performed by competitors. To get this information competition is referred to and
likewise clients of competitors are interviewed.
To benefit from this thesis not only the theoretical background will be looked upon, but
there will be approaches to completely integrate performance attribution.
Proposed usage will be in marketing where performance attribution is part of the
performance presentation. As it would be unwise not to use a controlling instrument for
its primary task the system will be used within controlling. With the help of
performance attribution, investment approaches shall be assessed and if necessary
improved or eliminated.
3
2 Implementation of a Performance Attribution System
The scope of this work is to implement a performance attribution system in a fund
management company. The effects of such a system upon marketing and controlling
shall be described and proposals for the choice of a system, its implementation and
organisational changes shall be given.
2.1 Definition of Problem
The reasons for banks to introduce third party investment funds are listed in the table
below. These reasons might serve as an indication for the choice of an asset manager,
too.
98
2000
2002
2003
Criteria
Ranking Ranking Ranking Quote Ranking Quote
Performance
30%
30%
Quality of Asset Manager and
Management
11
14%
13%
Reputation / Branding
12%
10%
Service
Investment Style and -Process
Low Volatility
Fees and Provisions
Other
10
Size and Liquidity
10
10
Funds Strategy and -Structure
10
Range of Products and Funds
17
12
12
11
Product Characteristics
12
13
12
Squeezing of Customers
14
13
Intensity of Relation
13
11
11
14
Legal and fiscal Factors
14
14
15
15
Figure 1: Criteria for the choice of Investment funds
1
During the last years, the performance of the funds and the quality of the management
became the most important factors to be regarded when an asset manager is chosen.
Performance attribution is used for detailed analysis and presentation of the
performance of a portfolio. Furthermore it demonstrates highly sophisticated
controlling mechanisms and therefore high quality of management.
1
Cp Busch in Westphal / Horstkotte / Ripper, Asset Management, p 44
4
To catch up with competitors or even to get a competitive advantage adequate
procedures and systems shall be integrated within the affected asset management
company.
The success of a portfolio manager and his strategy is mostly determined by comparing
the funds performance with the performance of a predefined benchmark
2
.
As a mere two figure performance comparison disregards risk and the development of
different parts of the investment, a more sophisticated approach is necessary.
This will be done by performance attribution which analyses the portfolio return
searching for its sources like a data mining software searches for undiscovered
connections in a data pool
3
.
2
Cp Günther, Stephan in Kleeberg / Rehkugler, Handbuch Portfoliomanagement, p 166 ff
3
Cp Numerical Algorithm group, http://www.nag.com/numeric/DR/DRdescription.asp
Cp Chapple, Mike, Data Mining an Introduction, http://databases.about.com/
library/weekly/aa100700a.htm
5
2.2 Relevance of Performance Attribution
The core activities in asset management are distribution, administration and portfolio
management.
Whereas administration and portfolio management might be done everywhere on earth
where a computer is available
4
distribution has to be handled on location.
The trend to establish a master KAG
5
to make use of economies of scale is growing.
Accounting, fund management, fund reporting, performance measurement and
controlling will be performed by such a master KAG. If a client is dissatisfied with the
funds in particular investment fields, there is no longer the need to transfer the complete
investment. As long as the client is satisfied with the KAG he might just change the
asset manager.
This trend enables foreign companies to enter the German market as there is no longer
the need to found their own KAG.
So the entrance barrier
6
for asset advisors from foreign countries consists of
establishing a distribution channel only, which leads to a highly competitive
environment where many entry strategy alternatives are used
7
.
Another reason for increasing competition is the overdue consolidation in the banking-
sector
8
. All these effects result in a "quiet revolution within the financial sector"
9
.
Besides deregulation and demographic change, the technological progress is the most
important driver of this revolution.
Hand in hand with the decreasing equity performance since 2000 the need of effective
controlling is growing.
4
Cp Hockmann in Westphal / Horstkotte / Ripper, Asset Management, pp 20 ff
5
Cp Busch in Westphal / Horstkotte / Ripper, Asset Management, p 46
6
Cp Kotler, Marketing Management, p 246
7
Cp Deresky, International Management, pp 242 ff
8
Cp Weber, Manfred, Bankmagazin für Führungskräfte der Finanzwirtschaft,July 2004, p
28
9
Cp Grothe / Spahn in Westphal / Horstkotte / Ripper, Asset Management, p 57
6
Figure 2: Value Chain for Asset Management
10
Within the affected asset management company the controlling mechanisms shall be
reinforced by a performance attribution system to strengthen the company's ability to
stand the increasing competition in the market.
2.3 Research Methods
To create this scientific thesis, research has been done. There are two sources of
research analysis of literature and a questionnaire sent to the relevant group,
institutional investors and asset managers.
2.3.1 Analysis of Literature
`... it is necessary to start by reviewing the literature carefully and in detail. In this case
however, the literature review might be used to identify and consider likely events...'
11
Performance attribution bases on portfolio management which is an established science.
Therefore literature is available and the relevant theories have to be mentioned.
Fortunately performance attribution itself is topic of scientific consideration, too. So it
is not necessary to develop new theories.
10
Cp Grothe / Spahn in Westphal / Horstkotte / Ripper, Asset Management, p 58
11
Cp Remenyi / Williams / Money / Swartz, Doing Research in Business and Management,
p 81
7
The existing theories have to be checked whether they might be used to solve the
problem and whether they have to be modified or not.
2.3.2 Questionnaire
Within this case the importance of the existing theories and the importance of the topic
itself for investors and asset managers are subject of the questionnaire.
The questionnaire has to be performed as the necessary knowledge is not available in
written or computerised form
12
.
Besides this some relevant information for asset managers might be drawn out of the
questionnaire, too
13
.
A detailed analysis about the questionnaire design will be given in chapter 4.4.
2.3.3 Interviews
Interviews with asset managers have not been performed in separate as after a few talks
with them it got obvious the results taken from interviews are similar to the results of
the questionnaire. As most of possible contacts arise from the same region and similar
companies, the danger of falsification because of overweighting their attitudes appeared
to be to big.
12
Cp Remenyi / Williams / Money / Swartz, Doing Research in Business and Management,
pp 150 ff
13
For example the relevance of personnel contact which is used to be compared with the
relevance of performance attribution, but delivers interesting outcomes by itself.
8
2.3.4 Conclusion
The research methods of this thesis follow the positivist approach to empirical
research
14
. The structure for this case is shown in the following figure:
Data Collection
Design
Problem
Literature Review
Primary Source
Literature
Control Source
Questionnaire
Developing
Solution
Implementing
Solution
Figure 3: Structure of Research Methods
14
Cp Remenyi / Williams / Money / Swartz, Doing Research in Business and Management,
p 83
9
3 Theory
According to Pieper there are three focal points for theories to handle while talking
about performance attribution. These are: measurement of performance, risk analysis
and performance attribution
15
.
As a scientific work should refer to all relevant theories, within this master thesis the
topics have been split up a little bit more.
It will start with theories for asset management, which are the portfolio theory or the
capital asset pricing model. Then it will go over to risk management followed by
performance measurement which is parted into general performance measurement and
risk adjusted performance measurement. Then performance contribution is introduced
followed by the topic of this thesis, performance attribution.
At last some relevant theories of marketing and controlling are mentioned as these
functional areas are treated, too.
3.1 Analysis of Relevant Theories
In this chapter all relevant theories are mentioned. As the range of mentioned theories is
broader then in the following chapter and as looking up of the sources shall be
simplified, the general sources are given in this chapter, even if they are referred to in
3.2. However, Figures, examples and literal citation will be mentioned in the particular
chapter.
3.1.1 Basic Principles of Asset Management
Many of the theories mentioned later will refer to some basics of security paper
management. Therefore the main theories of this should be mentioned.
3.1.1.1
Portfolio Theory according to Markowitz
The portfolio theory of Markowitz
16
postulates, that portfolio modelling is based upon
the two main factors risk and return of the individual securities. To avoid unsystematic
15
Cp Pieper; Handbuch Portfoliomanagement ( Articles combined by Kleeberg /
Rehkugler), 1998, p974
10
risk diversification is introduced. Efficient portfolios shall be created which means the
relation of risk to return is minimized. To calculate such portfolios the correlation of
security returns has an important role.
This model will be explained in more detail further on.
3.1.1.2
Capital Asset Pricing Model (CAPM)
Within the CAPM
17
the risky part of all portfolios is the same. Only the weight of the
risky part differs.
The most important result of the CAPM is the introduction of a beta factor which
describes whether the risk of a security is below or above market average.
This model will be described in detail, too.
3.1.1.3
Multi Beta CAPM
Within the multi beta CAPM
18
the return of a security depends on risk free return and a
lot of different beta factors. As this is `just' a modification of the CAPM this approach
will not be explained in detail.
3.1.1.4
Arbitrage Pricing Theory
The main assumptions of multi beta CAPM and APT are the same. But the APT waives
the knowledge of the market portfolio. As an enhancement a lot of micro- and macro-
economically risks are reviewed.
16
Cp Hielscher, Investmentanalyse,Oldenbourg Wien, pp 53 ff
Cp Sharpe / Alexander, Investments, pp 134 ff
Cp Shapiro / Balbirer, Modern Corporate Finance, pp189 ff
Cp Huang / Randall, Investment Analysis and Management, pp 447 ff
Cp Brealey / Myers, Principles of Corporate Finance, pp 187 ff
17
Cp Cuthbertson; Quantitative Financial Economics, pp 47 ff
Cp Hielscher, Investmentanalyse,Oldenbourg Wien, pp 77 ff
Cp Haugen, Modern Investment Theory, pp 197 ff
Cp Sharpe / Alexander, Investments, pp 194 ff
Cp Shapiro / Balbirer, Modern Corporate Finance, pp 196ff
Cp Brealey / Myers, Principles of Corporate Finance, pp 196 ff
Cp Loistl / Petag, Asset Management Standards, pp 116 ff
18
Cp Steiner, Bruns, Wertpapiermanagement, pp 27 ff
11
It is assumed, that all imbalances of the market will soon be erased by the acting of
arbitrageurs
19
. Further information may be taken from the named sources.
3.1.1.5
Market Model
It is the main assumption of this model
20
, that the risk is coming from one source only,
the market index. The return of securities is empirically determined and by regression
the model tries to give explanation of future returns.
3.1.1.6
Information Efficiency
The theory of information efficiency
21
describes the effects of markets which are in one
of three states of efficiency.
In a market with weak information efficiency all important data of the past are used to
determine a securities price. Fundamental analysis of securities might result in
advantage for investors.
Medium information efficiency means only the knowledge of insider information is of
use for an (unethical) investor. All other facts are included in the securities pricing.
At last strong information efficiency implements that all reasons for security pricing are
known by each member of the market and therefore are still priced in.
19
Cp Francis; Investments, 1991,pp 295 ff
Cp Reilly, Brown; Investment Analysis and Portfolio Management, pp 297 ff
Cp Shapiro / Balbirer, Modern Corporate Finance, pp 201 ff
Cp Cuthbertson; Quantitative Financial Economics, pp 61 ff
Cp Elton / Gruber, Modern Portfoliotheory and Investment Analysis, pp 368 ff
Cp Hielscher, Investmentanalyse,Oldenbourg Wien, pp 77 ff
Cp Haugen, Modern Investment Theory, pp 256 ff
Cp Sharpe / Alexander, Investments, pp 241 ff
Cp Huang / Randall, Investment Analysis and Management, pp 508 ff
20
Cp Hielscher, Investmentanalyse,Oldenbourg Wien, pp 70 ff
Cp Steiner, Bruns, Wertpapiermanagement, pp 37 ff
Cp Huang / Randall, Investment Analysis and Management, pp 497 ff
21
Cp Steiner, Bruns, Wertpapiermanagement, pp 40 ff
Cp Spencer, The Structure and Regulation of Financial Markets, p 1
Cp Breuer / Gürtler / Schuhmacher, Portfoliomanagement, pp 375 ff
12
3.1.2 Risk Measurement
As portfolio managers always try to create efficient portfolios according to CAPM,
there is the need to specify the certain risk of an investment.
3.1.2.1
Volatility
To use volatility of a portfolio return as a ratio for its risk is quite a common
approach
22
. It is described later on.
3.1.2.2
Value at Risk
The Value at Risk (VaR) approach is another statistical approach. The VaR describes
the maximal loss for an investment at a certain possibility within a certain time frame.
The value of the VaR is estimated by certain stochastic simulations (like Monte Carlo
methods) or by a parametric approach. One example for the parametric approach is to
determine the VaR of a portfolio with the help of covariance matrix of its
constituents
23
.
3.1.2.3
Probability of Failure
This is a rather evident risk ratio
24
. It estimates the probability of a certain loss within a
certain time frame. The basic idea is, that portfolio returns are Gauss variables. The
mathematical approach is similar to the VaR as this approach takes just a little different
point of view upon the same foundations.
3.1.2.4
Asymmetric Risk Values
As portfolio returns often can not be seen as Gauss variables this approach uses
asymmetrical distribution within its calculations
25
.
22
Cp Fischer, Performanceanalyse in der Praxis ,p 233
Cp Steiner, Bruns, Wertpapiermanagement, pp 57
23
Cp JP Morgan / Reuters 1996
Cp Fischer, Performanceanalyse in der Praxis ,p 246
24
Cp Fischer, Performanceanalyse in der Praxis ,p 251
Cp Steiner, Bruns, Wertpapiermanagement, pp 62 ff
25
Cp Fischer, Performanceanalyse in der Praxis ,p 253
13
The main idea is to specify a risk level to fall short of a certain pre-defined
performance.
3.1.2.5
Duration
The duration of a bond describes how much the return is changing if the yield changes.
So it describes the risk of bond portfolios in case of yield changes
26
.
3.1.3 Performance Measurement
Within performance attribution the performance or the return of portfolios,
benchmarks, sectors, equities or bonds is compared to each other. Therefore it is of
great importance to know how such performance is calculated.
Within performance attribution return and performance are widely used synonymously
as in this case the one-dimensional value is regarded. Within this thesis the concept of
Performance as risk weighted return
27
is not considered in detail. However, the basic
concept has to be explained to create an idea in how far this approach differs from the
concept of performance used for performance attribution.
3.1.3.1
Performance as Return
Within this chapter some methods of estimating performance will be mentioned.
Starting with the money weighted rate of return and the time weighted rate of return the
explanations will go on with the BVI method and the modified Dietz method.
The Dietz method itself, which is a rather simple approach, will not be explained. Nor
will the BAI method which is quite similar to the Dietz method.
26
Cp Fischer, Performanceanalyse in der Praxis ,p 240
Cp Elton/ Gruber, Modern PortfolioTheory and Investment Analysis, pp 534 ff
Cp Steiner, Bruns, Wertpapiermanagement, pp 155 ff
Cp Hull, Futures and Options, pp 155 ff
Cp Brealey / Myers, Principles of Corporate Finance, pp 674 ff
27
Cp Steiner, Bruns, Wertpapiermanagement, p 568
14
As the calculated figures are the basis for performance attribution, the proper choice of
the return calculation is of great importance for the quality of performance attribution
results later on
28
.
3.1.3.1.1
Return Estimation of Securities
The identification of equity return seems quite easy. It is just the difference between the
buying and the selling price.
If dividend is paid within the calculation period it is a little bit trickier, as the return
figure has to be adjusted. If more than one dividend payments are done within the
calculation period a BVI like calculation has to be used.
At last the exchange rate has to be regarded, too if the investment is done in another
than the home currency.
For a bond investment the situation is much worse
29
. There are a lot of market practices
how the bond return might be calculated.
Starting with computation of accrued interest, where the days might be taken as actual
days or 30 for the days of the month going to actual day for a year over 365 or 360.
Almost all combinations have been used sometimes, somewhere. Nowadays the ISMA
calculation (actual days (month) / actual days (year)) is used mostly.
Furthermore return figures a harder to calculate as accrued interest is always part of the
deal. Many methods according to Braeß / Fangmeyer, West LB, Moosmüller or ISMA
are available.
As portfolio evaluation is the focus of this thesis, all the mentioned methods are not
referred to.
3.1.3.1.2
Money weighted Rate of Return
This methodology tries to evaluate the average interest of an investment
30
.
28
Cp Pieper; Handbuch Portfoliomanagement ( Articles combined by Kleeberg /
Rehkugler), 1998, p975
29
Cp Eller, Modernes Bondmanagement, pp12 ff
30
Cp Fischer, Performanceanalyse in der Praxis ,p 11
Cp Huang / Randall, Investment Analysis and Management, pp 512 ff
Cp Hielscher, Investmentanalyse, pp 40ff
15
3.1.3.1.3
Time weighted Rate of Return
The estimation of an interest rate without influence of cash flow is the main idea of this
approach
31
.
3.1.3.1.4
BVI Method
This approach regards the compound interest of investments, especially security
funds
32
.
3.1.3.1.5
Modified Dietz Method
This approach also regards compound interest, but delivers an average interest like the
money weighted return as a result
33
.
3.1.3.1.6
Dietz Method
To estimate a return all cash flow is assumed to take place in the middle of a period.
This method is an evaluation of interest. This is the linear approximation towards the
BAI method, which uses the same assumptions and a quadratic equation to solve this
problem
34
.
3.1.3.2
Performance as Risk weighted Return
While the performance estimation methods mentioned in 3.1.3.1 always deliver an
absolute return value as performance, within this chapter performance is seen relatively
to the risk which has been taken by the asset advisor to realize this performance. The
first four methodologies will be described briefly in 3.2.
31
Cp Fischer, Performanceanalyse in der Praxis ,p 15
Cp Huang / Randall, Investment Analysis and Management, pp 514 ff
Cp Hielscher, Investmentanalyse, pp 40ff
32
Cp Fischer, Performanceanalyse in der Praxis ,p 20
Cp Albrecht / Maurer, Investment- und Risikomanagement, pp 66ff
33
Cp Fischer, Performanceanalyse in der Praxis ,p 40
34
Cp Fischer, Performanceanalyse in der Praxis ,p 46
16
Besides the tracking error all ratios follow the same principle. They take an excess
return and divide it by a risk ratio. Only the kind of excess return and risk ratio differ
from case to case.
3.1.3.2.1
Sharpe Ratio
The Sharpe ratio takes the excess return of the portfolio over the benchmark and
divides it by the volatility of the portfolio return. The delivered ratio might be used to
compare different portfolios or benchmarks
35
.
3.1.3.2.2
Treynor Ratio
The Treynor ratio is determined quite similar. Instead of the volatility the factor of the
CAPM (see chapter 3.2.2) is used as the risk component
36
.
3.1.3.2.3
Jensen Ratio
The Jensen is an absolute figure which describes the possibility of an asset advisor to
identify underestimated stocks. Dividing it by the volatility of the portfolio return
delivers the Treynor / Black ratio which again delivers return / risk ratio
37
.
3.1.3.2.4
Tracking Error
As portfolios consist of fewer securities than the benchmark, the tracking error is used
to deliver a ratio to assess the quality of this sampling
38
.
35
Cp Fischer, Performanceanalyse in der Praxis ,p 271
Cp Huang / Randall, Investment Analysis and Management, pp 517 ff
Cp Haugen, Modern Investment Theory, pp 288 ff
Cp Steiner, Bruns, Wertpapiermanagement, pp 576 ff
36
Cp Fischer, Performanceanalyse in der Praxis ,p 273
Cp Haugen, Modern Investment Theory, pp 287 ff
Cp Steiner, Bruns, Wertpapiermanagement, pp 579 ff
Cp Huang / Randall, Investment Analysis and Management, pp 518 ff
37
Cp Fischer, Performanceanalyse in der Praxis ,p 275
Cp Haugen, Modern Investment Theory, pp 284 ff
Cp Steiner, Bruns, Wertpapiermanagement, pp 581 ff
Cp Huang / Randall, Investment Analysis and Management, pp 519 ff
38
Cp Fischer, Performanceanalyse in der Praxis ,p 257
Cp Steiner, Bruns, Wertpapiermanagement, pp 70 ff
17
3.1.3.2.5
Differential Return
As the Jensen uses the factor of the CAPM to be calculated, the differential return
uses volatility, keeping the rest of the procedure the same
39
.
3.1.3.2.6
Risk Adjusted Performance
With Sharpe- and Treynor- ratio a ranking of portfolios according to their risk might be
performed. Jensen alpha and differential return try to compare absolute return of a
portfolio
40
.
The risk adjusted return has been developed to refer to both aims using one ratio.
3.1.3.2.7
Information Ratio
The information ratio is the excess return of a portfolio divided by the tracking error as
the risk part. So it is similar to Sharpe and Treynor ratio
41
.
3.1.3.2.8
Sortino Ratio
The Sortino ratio again divides an excess return, here the difference between the return
to the minimum return according to the asymmetric risk valuation, and divides it by the
deviation risk descended from the same approach
42
.
3.1.4 Performance Contribution
Performance contribution
43
is the mathematic which describes which weighting and
which security is responsible for the development of the portfolio.
The difference to performance attribution is that no comparison to a benchmark is done.
Performance contribution is covered in detail in 3.2.7.
39
Cp Fischer, Performanceanalyse in der Praxis ,p 278
40
Cp Fischer, Performanceanalyse in der Praxis ,p 280
41
Cp Fischer, Performanceanalyse in der Praxis ,p 283
Cp Farrell, Portfolio Management, p. 544
42
Cp Fischer, Performanceanalyse in der Praxis ,p 284
43
Cp David Spaulding, Investment Performance Attribution, p 15
18
3.1.5 Performance Attribution
Within performance attribution the treatment of the equity side is well done. Some
models exist mostly based upon a Brinson approach, which has been more or less
enhanced. In regard to the case the considered theories are those, which are
implemented in the popular performance attribution systems.
Where this is not a problem on the equity side, for bonds the situation is a little bit more
complicated as the methods to perform an attribution are by far not standardized, which
is noticeable by the great number of approaches. Finally an approach is chosen which
fits best the HTCM investment process and can be combined with the elected
performance attribution system.
Even some vendors of performance attribution systems include handling of derivatives
in their systems their treatment is marginally, as the practical relevance in HTCM is
low.
3.1.5.1
Brinson Hood Beebower
This is the classic approach
44
of equity performance attribution. As it is the standard
approach it will be described briefly later on.
3.1.5.2
Brinson / Fachler
The Brinson Fachler approach
45
is an improvement of the BHB model. It also will be
described briefly later on.
3.1.5.3
Karnosky Singer
The Karnosky Singer approach
46
enhances the BHB model in regard to currency
exchange variation. As it is rather complex and as it looses relevance because of the
European Currency Union it will not be described in detail.
44
Cp David Spaulding, Investment Performance Attribution, pp 31 ff
Cp Schwyser's Study Notes, pp 145 ff
Cp Sharpe, The Journal of Portfoliomanagement, 1992, pp 16 ff
45
Cp Brinson / Fachler, The Journal of Portfolio Management, July/August 1985, pp 73
76
Cp David Spaulding, Investment Performance Attribution, pp 42 ff
46
Cp Singer / Karnosky, The Journal of Portfoliomanagement,Volume 21 1995, pp84 - 92
19
3.1.5.4
Geometric Approach
Having harder mathematics and a clearness which does not convince at first view, the
geometric approach
47
offers more elegant solutions and a lot of advantages compared
with the arithmetic models mentioned above. Therefore it too will be described briefly.
3.1.5.5
Fixed Income Attribution
As there has no standard been established yet for fixed income attribution
48
please refer
to the footnotes to check the different literature. Within this Master Thesis the approach
implemented in the Wilshire system will be discussed.
3.1.5.6
Linking across Time
The results of performance attribution from a given period A to B and the results of B
to C should equal the direct calculation for the time frame A to C. As this is not as easy
as it sounds
49
, this topic will be discussed for geometric and arithmetic attribution
separately.
3.1.6 Fund Management
3.1.6.1
Active / Passive Management
The differences between passive management, which means replicating a benchmark
and active management, which is making profit from information deficits of the market
are explained in more detail later on.
Cp Allen Gregory, The Journal of Portfoliomanagement, Volume 18 1991, pp 59 - 65
Cp David Spaulding, Investment Performance Attribution, pp 87 ff
47
Cp David Spaulding, Investment Performance Attribution, pp 105 ff
48
Cp David Spaulding, Investment Performance Attribution, pp 53 ff
49
Cp Laker, The Journal of Portfolio Management, Summer 2002
Cp Spaulding, Journal of Performance Measurement, Spring 2002, pp 32 39
Cp Frongello, Journal of Performance Measurement, Fall 2002, pp 54 - 67
20
3.1.6.2
Other Management Styles
There are different styles how to invest in assets. Some examples are top down versus
bottom up, value versus growth or large cap versus mid cap
50
. More details might be
gathered from the named sources.
3.2 Selection of Appropriate Theories
Within this chapter the appropriate theories will be explained in detail.
3.2.1 Portfolio Selection according to Markowitz
The portfolio selection theory
51
is based upon an empiric observation of equity market
participants. As investors diversify their investments, it is not their motivation to get a
high return, in this case they would have to determine the best performing equity to
invest all their money, but there is an appreciation of values between risk and return.
A formula to take these both motivations into consideration is according to Markowitz:
=
=
n
i
i
i
P
R
x
R
1
*
R
P
= expected portfolio return
x
i
= weight of element
R
i
= return of individual component
i = number of element
n = amount of elements
The risk of such an investment is calculated by the empirical variance (
2
) or the
deviation () of the portfolio return. The formula for the deviation is:
50
Cp Behrenwaldt in Leser / Rudolf, Handbuch Institutionelles Asset Management, pp 324
ff
51
Cp Hielscher, Investmentanalyse,Oldenbourg Wien, pp 53 ff
Cp Sharpe / Alexander, Investments, pp 134 ff
Cp Shapiro / Balbirer, Modern Corporate Finance, pp189 ff
Cp Huang / Randall, Investment Analysis and Management, pp 447 ff
Cp Brealey / Myers, Principles of Corporate Finance, pp 187 ff
Cp Steiner, Bruns, Wertpapiermanagement, pp 7 ff
Cp Reilly, Brown; Investment Analysis and Portfolio Management, pp 277 ff
Details
- Seiten
- Erscheinungsform
- Originalausgabe
- Jahr
- 2004
- ISBN (eBook)
- 9783832486693
- ISBN (Paperback)
- 9783838686691
- DOI
- 10.3239/9783832486693
- Dateigröße
- 1.2 MB
- Sprache
- Englisch
- Institution / Hochschule
- Hogeschool Zeeland – unbekannt
- Erscheinungsdatum
- 2005 (März)
- Note
- 1,0
- Schlagworte
- risk return performance measurement bank financials
