Copulas are mathematical objects that fully capture the dependence structure among
random variables and hence, offer a great flexibility in building multivariate stochastic models. Since their introduction in the early 50s, copulas have gained a lot of
popularity in several fields of applied mathematics, like finance, insurance and reliability theory. Nowadays, they represent a well-recognized tool for market and
credit models, aggregation of risks, portfolio selection, etc.
Moreover, such a large interest in the applications of copulas have spurred researchers and scientists in investigating and developing new theoretical methods
and tools for handling uncertainty in practical situations.
The workshop on “Copula Theory and Its Applications”, which took place in
Warsaw (Poland) on 25th–26th September 2009, represented a good opportunity
for intensive exchange of ideas about recent developments and achievements that
can contribute to the general development of the field. The talks presented at this
event have focused on several interesting theoretical problems as well as empirical
applications, especially to finance, insurance and reliability.
In order to make all these contributions available to a larger audience, we have
planned to prepare a volume collecting selected contributions of the workshops and
several original survey papers about important aspects of copula theory and its applications.
The book is divided into two main parts: Part I Surveys contains 11 manuscripts
giving an up-to-date account of some aspects of copula models. Part II Contributions
collects an extended version of 6 talks selected from presented at the workshop in
Warsaw.
Our special thanks go to the authors for their willingness to contribute to this volume, and to our colleagues Erich Peter Klement, Radko Mesiar, José Juan Quesada
Molina, Carlo Sempi, and Fabio Spizzichino, who, as members of the Scientific
Committee of the workshop, contributed to the scientific success of this event.
Every paper has been submitted to, at least, one referee: we want to thank all of
them for their collaboration.
The professional work of the Organizing Committee was greatly appreciated, as
well as the support of the co-sponsors of this conference.
Finally, we are indebted to our publisher Springer Verlag and prepublisher
Integra Software Services, in particular to Dr. Niels Peter Thomas, Alice Blanck
and Sharmila Krishnamurthy for their assistance in the editorial process.
Linz, Austria Fabrizio Durante
Berlin, Germany Wolfgang Härdle
Warszawa, Poland Piotr Jaworski
Toru′ n, Poland Tomasz Rychlik
January 2010
Part I Surveys
1 Copula Theory: An Introduction .............................. 3
Fabrizio Durante and Carlo Sempi
1.1 Historical Introduction . . ................................... 3
1.1.1 Outline .......................................... 6
1.2 Preliminaries on Random Variables and Distribution Functions . . . 6
1.3 Copulas: Definitions and Basic Properties . . . . ................. 9
1.4 Sklar’s Theorem .......................................... 12
1.5 Copulas and Random Vectors . . . ............................ 14
1.6 Families of Copulas . . . . ................................... 15
1.6.1 Elliptical Copulas . . . . . ............................ 16
1.6.2 Archimedean Copulas . . ............................ 17
1.6.3 EFGM Copulas ................................... 19
1.7 Constructions of Copulas ................................... 20
1.7.1 Copulas with Given Lower Dimensional Marginals..... 20
1.7.2 Copula-to-Copula Transformations . . ................. 21
1.7.3 Geometric Constructions of Copulas . ................. 22
1.8 Copula Theory: What’s the Future? . . ........................ 23
References ..................................................... 24
2 Dynamic Modeling of Dependence in Finance via Copulae Between
Stochastic Processes......................................... 33
Tomasz R. Bielecki, Jacek Jakubowski and Mariusz Niew˛egłowski
2.1 Introduction .............................................. 33
2.2 Lévy Copulae . . .......................................... 35
2.3 Semimartingale Copulae ................................... 39
2.3.1 Copulae for Special Semimartingales ................. 39
2.3.2 Consistent Semimartingale Copulae . ................. 48
2.4 Markov Copulae .......................................... 54
2.4.1 ConsistentMarkovProcesses........................ 55
ix
x Contents
2.4.2 Markov Copulae: Generator Approach . . . ............. 57
2.4.3 Markov Copulae: Symbolic Approach................ 63
2.5 Applications in Finance . ................................... 69
2.5.1 Pricing Rating-Triggered Step-Up Bonds via Simulation . 70
2.5.2 Model Calibration and Pricing....................... 72
References ..................................................... 75
3 Copula Estimation .......................................... 77
Barbara Choro′ s, Rustam Ibragimov and Elena Permiakova
3.1 Introduction .............................................. 77
3.2 Copula Estimation: Random Samples with Dependent Marginals . 78
3.2.1 Parametric Models: Maximum Likelihood Methods
and Inference from Likelihoods for Margins . . . . . ...... 78
3.2.2 SemiparametricEstimation ......................... 80
3.2.3 Nonparametric Inference and Empirical Copula Processes 81
3.3 Copula-Based Time Series and Their Estimation . . ............. 82
3.3.1 Copula-Based Characterizations for (Higher-Order)
MarkovProcesses ................................. 82
3.3.2 Parametric and Semiparametric Copula Estimation
Methods for Markov Processes . . . . . ................. 83
3.3.3 Nonparametric Copula Inference for Time Series . ...... 84
3.3.4 Dependence Properties of Copula-Based Time Series . . . 85
3.4 Further Copula Inference Methods........................... 86
3.5 EmpiricalApplications..................................... 87
References ..................................................... 89
4 Pair-Copula Constructions of Multivariate Copulas............... 93
Claudia Czado
4.1 Introduction .............................................. 93
4.2 Pair Copula Constructions of D-Vine, Canonical and Regular
VineDistributions......................................... 94
4.2.1 Pair-Copula Constructions of D-Vine and Canonical
VineDistributions ................................. 94
4.2.2 Regular Vines Distributions and Copulas . ............. 96
4.3 Estimation Methods for Regular Vine Copulas .................100
4.4 Model Selection Among Vine Specifications . .................103
4.5 ApplicationsofVineDistributions ...........................105
4.6 SummaryandOpenProblems...............................106
References .....................................................107
5 Risk Aggregation............................................111
Paul Embrechts and Giovanni Puccetti
5.1 MotivationsandPreliminaries...............................112
5.1.1 TheMathematicalFramework.......................112
5.2 Bounds for Functions of Risks: The Coupling-Dual Approach . . . . 113
5.2.1 Application 1: Bounding Value-at-Risk . . .............115
Contents xi
5.2.2 Application 2: Supermodular Functions . . .............119
5.3 TheCalculationoftheDistributionoftheSumofRisks .........120
5.3.1 OpenProblems ...................................124
References .....................................................125
6 Extreme-Value Copulas......................................127
Gordon Gudendorf and Johan Segers
6.1 Introduction ..............................................127
6.2 Foundations ..............................................128
6.3 Parametric Models . . . . . ...................................131
6.3.1 Logistic Model or Gumbel–Hougaard Copula . . . . ......132
6.3.2 Negative Logistic Model or Galambos Copula . . . ......132
6.3.3 Hüsler–Reiss Model . . . ............................133
6.3.4 The t-EV Copula..................................134
6.4 Dependence Coefficients ...................................134
6.5 Estimation ...............................................136
6.5.1 ParametricEstimation..............................137
6.5.2 Nonparametric Estimation . . ........................138
6.6 Further Reading ..........................................140
References .....................................................141
7 Construction and Sampling of Nested Archimedean Copulas.......147
Marius Hofert
7.1 Introduction ..............................................147
7.2 Nested Archimedean Copulas . . . ............................149
7.3 A Sufficient Nesting Condition . . ............................151
7.4 Construction of Nested Archimedean Copulas .................153
7.5 Sampling Nested Archimedean Copulas . . . . . .................155
7.6 Conclusion . ..............................................159
References .....................................................159
8 Tail Behaviour of Copulas....................................161
Piotr Jaworski
8.1 Introduction ..............................................161
8.2 Tail Expansions of Copulas . . . . ............................163
8.2.1 Characterization and Properties of Leading Parts . ......167
8.2.2 Relatively Invariant Measures on[0,∞)
n
..............168
8.3 Examples of Tail Expansions . . . ............................169
8.3.1 Homogeneous Copulas . ............................169
8.3.2 Diagonal Copulas . . . . . ............................169
8.3.3 Absolutely Continuous Copulas . . . . .................171
8.3.4 Archimedean Copulas . . ............................172
8.3.5 Multivariate Extreme Value Copulas . .................177
8.4 Applications .............................................178
8.4.1 Tail Conditional Copulas...........................178
8.4.2 Extreme Value Copulas of a Given Copula ............180
xii Contents
8.4.3 Regularly Varying Random Vectors with a Given Copula 181
8.4.4 ValueatRisk .....................................182
References .....................................................185
9 Copulae in Reliability Theory (Order Statistics, Coherent Systems) . 187
Tomasz Rychlik
9.1 Coherent Systems.........................................187
9.2 Signatures . ..............................................189
9.2.1 Components with i.i.d. Lifetimes . . . .................189
9.2.2 MixedSystems ...................................190
9.2.3 Components with Exchangeable Lifetimes ............192
9.3 Bounds for Exchangeable Lifetime Components . . .............194
9.3.1 Distribution Bounds . . . ............................194
9.3.2 Expectation Bounds . . . ............................196
9.4 Characterizations ofk-Out-of-nSystem Lifetime Distributions . . . 198
9.4.1 General Copula Joint Distribution....................199
9.4.2 Absolute Continuous Copula Joint Distribution . . . ......200
9.4.3 Variance Bounds..................................203
9.5 FinalRemarks............................................205
References .....................................................206
............................................321