Assessing Risk Assessment: Towards Alternative Risk Measures for Complex Financial Systems

377 Pages·2017·2.39 MB·English

by  Christian Hugo Hoffmann (auth.)

Christian Hugo Hoffmann
Assessing Risk
Assessment
Towards Alternative Risk Measures
for Complex Financial Systems
Assessing Risk Assessment
Christian Hugo Hoffmann
Assessing Risk
Assessment
Towards Alternative Risk Measures
for Complex Financial Systems
Christian Hugo Hoffmann
Kreuzlingen, Switzerland
Dissertation der Universität St. Gallen, 2017
Die Universität St. Gallen, Hochschule für Wirtschafts-, Rechts- und Sozialwissen-
schaften sowie internationale Beziehungen (HSG), gestattet hiermit die Drucklegung
der vorliegenden Dissertation, ohne damit zu den darin angesprochenen Anschauun-
gen Stellung zu nehmen.
OnlinePlus material to this book is available on
http://www.springer.com/978-3-658-20032-9
ISBN 978-3-658-20031-2 ISBN 978-3-658-20032-9 (eBook)
https://doi.org/10.1007/978-3-658-20032-9
Library of Congress Control Number: 2017957211
Springer Gabler
© Springer Fachmedien Wiesbaden GmbH 2017
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole
or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher
nor the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publisher remains
neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Printed on acid-free paper
This Springer Gabler imprint is published by Springer Nature
The registered company is Springer Fachmedien Wiesbaden GmbH
The registered company address is: Abraham-Lincoln-Str. 46, 65189 Wiesbaden, Germany
Table of Contents

List of Tables and Figures ................................................................................................................ ix
Abstract ............................................................................................................................................. xi

Zusammenfassung .......................................................................................................................... xiii

Introduction ....................................................................................................................................... 1
Part I: Concepts, Model Level and Risk Assessment ................................................................ 21
1. Introduction to Part I ................................................................................................................... 23
2. Literature Synthesis, Theoretical Background and Research Focus ............................................ 25

2.1. Complexity and Modern Financial Systems ...................................................................... 25

2.2. Risk and Risk Management in the Financial World .......................................................... 32

2.2.1. Risk modeling ..................................................................................................... 37

2.2.2. Value at Risk (VaR) ............................................................................................ 41

2.2.3. Expected Shortfall (ES) ...................................................................................... 47

2.3. Systemic Risk Assessment ................................................................................................ 49

2.3.1. Tools primarily for regulators:

Conditional Value at Risk (CoVaR) and Systemic Expected Shortfall (SES) ..... 51

2.3.2. Extreme Value Theory (EVT) ............................................................................. 53

2.4. General Appraisal ............................................................................................................. 55

2.4.1. Advantages of conventional risk models and measures....................................... 56

2.4.2. Weaknesses of conventional risk models and measures ...................................... 58

2.5. Excursus: Benoît Mandelbrot's Plea for Fractal Methods .................................................. 63
3. Research Questions..................................................................................................................... 73
4. On an Adequate Concept of Risk and Systemic Risk in the Realm of Banking ........................ 75

4.1. The Notion of Risk ........................................................................................................... 75

4.2. The Concept of Systemic Risk .......................................................................................... 91
5. On the Relevance of Systemic Risks for Banks ........................................................................ 103

5.1. Why should Banks take account of, and try to deal with, Systemic Risks? ..................... 103

5.2. What are concrete Systemic Risk Scenarios for Banks? ................................................. 115
6. Dealing with Quantitative Risk Management in Banking as a Complex Systems Problem ...... 123

6.1. A Trichotomy of Scientific Problems – Warren Weaver’s Scheme as a General

Answer to How to Manage Complexity .......................................................................... 128

6.1.1. Tackling disorganized complexity versus organized simplicity ........................ 128

6.1.2. Disorganized complexity and statistical techniques. ......................................... 131

6.1.3 Tackling organized complexity: open questions remain....................................... .134

6.1.4. Synopsis ............................................................................................................ 136

6.2. Weaver’s Taxonomy Revisited: Attempts of Clarification, Extension and Refinement .. 139

6.2.1. Approaches towards the operationalization of Weaver’s concept of organized

complexity ........................................................................................................ 139

6.2.2. The bigger picture of complexity and randomness ............................................ 142
vi Table of Contents

6.3. Organized Complexity, Financial Systems and Assessing Extreme and Systemic Risks 152

6.3.1. On the level of structures .................................................................................. 154

6.3.2. On the level of events........................................................................................ 155

6.4. A Tentative Bottom Line ................................................................................................ 157
7. The Fundamental Inadequacy of Probability Theory as a Foundation for Modeling Systemic

and Extreme Risk in a Banking Context ................................................................................... 159

7.1. Philosophical Roots of the Problem of Induction: some Preliminaries ............................... 161

7.2. Probability Theory in a Nutshell, its Embeddedness and its Applications....................... 164

7.3. The Central Argument against using Probability Theory for Financial Risk

Management ................................................................................................................... 175

7.4. Linking the Central Argument with the Current State of the Literature (IIIa)-c)) ........... 183
8. Conclusion to Part I .................................................................................................................. 187

8.1. Résumé................................................................................................................................... 188

8.2. Outlook: Explanatory Models for In-House Risk Management in Banking .................... 191
Part II: The Transition to the Decision Level, Risk Assessment and Management ............... 195
9. Introduction to Part II ............................................................................................................... 197
10. The Critical Turn: The Renaissance of Practical Wisdom ........................................................ 199
11. Scenario Planning in a Nutshell and its Role in Risk Management in Banking ........................ 205
12. Strengths and Weaknesses of Scenario Planning as a Risk Management Tool ......................... 213
13. Deriving Lessons for Rethinking the Approach to Assessing Extreme and Systemic Risks ..... 219
Part III: In Search of a New Paradigm:

The Third Way as a Road to Logic-Based Risk Modeling (LBR) ............................. 223
14. Introduction to Part III .............................................................................................................. 225
15. Theoretical Foundations of a Logic-Based Risk Modeling (LBR) Approach ........................... 231

15.1. A less Restrictive Axiomatization ................................................................................... 231

15.2. Non-Probabilistic Models of Uncertainty ....................................................................... 239

15.3. Ranking Theory .............................................................................................................. 243

15.4. Syntax of a Language for Describing Contracts and Correlations ................................... 246

15.5. Semantics: Financial Contracts as Uncertain Sequences in a Non-Probabilistic

Risk Model Context ....................................................................................................... 252

15.5.1. Uncertain sequences by example ...................................................................... 253

15.5.2. From contract value to risk ............................................................................... 257

15.5.3. Formalization of the approach ........................................................................... 258

15.5.4. Concrete instantiations of uncertainty monads: ranking functions .................... 265

15.5.5. Evaluating risk models ...................................................................................... 270

15.6. Model Interpretation and Output: An Exact, Explanatory Scenario Planning Method .... 275
16. Case Study: LTCM and Extreme Risk ...................................................................................... 279

16.1. Example Trade ................................................................................................................ 280

16.2. A Fixed Income Portfolio in LBR ................................................................................... 281
16.3. Analysis .......................................................................................................................... 284

16.3.1. Overview .......................................................................................................... 285

16.3.2. Zoom and filter ................................................................................................. 286

16.3.3. Details on demand............................................................................................. 288

16.4. Discussion and Conclusion ............................................................................................. 288
Table of Contents vii
17. Managerial Implications ........................................................................................................... 291
18. Scales of Measurement and Qualitative Probabilities ............................................................... 297
19. Model Validation ...................................................................................................................... 303
Part IV: Meta Level: Thinking about Thinking and Practices – What it Means to Reach
Effective Risk Management Decisions ......................................................................... 321
20. Introduction to Part IV as Overall Conclusion .......................................................................... 323
21. Escaping the Traps for Logicians:
Towards Decision-Making Competency in Risk Management ................................................. 325
22. Final Remarks and a Path for Future Research ......................................................................... 339
References ...................................................................................................................................... 345

List of Appendices............................................................................................................................... 377
To access the book’s appendix, please visit www.springer.com and search for the author’s
name. URL: http://www.springer.com/978-3-658-20032-9
List of Tables and Figures
Table 1: Methods and rules to realize the research objectives. ..................................................... 16
Table 2: Risks banks face and how they measure and manage them. ........................................... 42
Table 3: Overview of systemic risk models and how they might be classified. ............................ 57
Table 4: Classification system for risk definitions and characterization of different risk
definition categories. ...................................................................................................... 78
Table 5: A suggested taxonomy of uncertainties and complexities. ........................................... 138
Table 6: A critic's comparison of quantitative risk modeling (theme of Part I) and
qualitative risk analyses (theme of Part II). .................................................................. 203
Table 7: Strengths and weaknesses of scenarios and scenario analysis. ..................................... 214
Table 8: Summary of the vocabulary for financial contracts. ..................................................... 252
Table 9: Ranking function , showing the likelihood of default (di) for three counterparties
A, B and C. The most likely case is , meaning no defaults. ................. 266
Table 10: obtained by co𝓀𝓀n𝑓𝑓ditioning .
(Table 10 is obtained by applying D𝑓𝑓ef𝐴𝐴. 1∧5 .𝑓𝑓11𝐵. ∧to 𝑓𝑓th𝐶𝐶e ranking function in Table 9.) ...... 269
Table 11: 𝓀𝓀Ra𝑓𝑓t′es for LTCM trade. ............𝓀𝓀...𝑓𝑓.. .𝑒𝑒...𝑖𝑖.𝑡𝑡..ℎ.. .𝑓𝑓..𝐴𝐴... ..∨... .𝑓𝑓..𝐵......................................................... 281
Table 12: Validating LBR risk models along a collective of reasonable criteria. ......................... 306

Figure 1: The systems view. .......................................................................................................... 27
Figure 2: The disassembly of complexity I. ................................................................................... 33
Figure 3: Images of three different fractals. ................................................................................... 66
Figure 4: Two relevant risk concepts: Risk I encompasses Risk II and uncertainty. ...................... 89
Figure 5: Risk sharing and systemic risks I. ................................................................................ 107
Figure 6: Risk sharing and systemic risks II. ............................................................................... 110
Figure 7: Event-oriented view of the world. ................................................................................ 111
Figure 8: The feedback view of the world, Part I......................................................................... 111
Figure 9: The feedback view of the world, Part II. ...................................................................... 112
Figure 10: Gross value of $1 invested in LTCM, March 1994 – September 1998. ........................ 119
Figure 11: The evolution of the Systems Approach. ...................................................................... 125
Figure 12: The line of argumentation in Chapter 6 and its embeddedness. .................................... 127
Figure 13: Weaver on probability theory, statistics and fields of application ................................ 133
Figure 14: The disassembly of complexity II: The extended framework. ...................................... 137
Figure 15: The bigger picture of complexity and randomness. ...................................................... 147
Figure 16: Bringing complexity, randomness and risk together..................................................... 151
Figure 17: Heavy/fat and long versus thin and short tails. ............................................................... 171
Figure 18: Normal distributions are not the (new) norm(al)............................................................. 171
Figure 19: The modeling relations. ................................................................................................ 192
Figure 20: Scenario planning as a circular and iterative process from a feedback view
of the world. ................................................................................................................. 209
Figure 21: The solid house of LBR. .............................................................................................. 232
Figure 22: Transactions of an unsecured loan annotated with uncertain values. ............................ 254
Figure 23: Transactions in a bond brought forward as collateral for a loan. .................................. 256
Figure 24: Transactions in a collateralized loan. ........................................................................... 257
x List of Tables and Figures
Figure 25: Two commutative diagrams. ........................................................................................ 261
Figure 26: Transactions in a collateralized loan, annotated with default information. ................... 269
Figure 27: Risk model cl', showing the collateralized loan after applying the scenario from
Example 15.14. ............................................................................................................ 275
Figure 28: Repo rate rp expressed as an uncertain sequence. ........................................................ 282
Figure 29: Libor rate lbr expressed as an uncertain sequence. ....................................................... 283
Figure 30: Part I of LTCM Trade. ................................................................................................. 283
Figure 31: Part II of LTCM Trade. ................................................................................................ 284
Figure 32: Range of values over time. ........................................................................................... 286
Figure 33: Range of values in the time period from one to eight days. .......................................... 287
Figure 34: Scenario s modeling the default of A, B or C on day two or three. .............................. 288
1
Figure 35: The disassembly of complexity III: The unifying framework. ..................................... 299
Figure 36: The notion of qualitative probability as head of a cluster. ............................................ 301
Abstract
Traditional and much of modern risk assessment and management is old fash-
ioned, unrealistic and trapped in a dogmatic slumber. It generates its own risks,
evoked by an improper notion of risk which lacks a systemic viewpoint. A cul-
ture dominates theory (economics and finance) and practice (financial institu-
tions) where risk modeling is no longer a functional tool but has become an end
in itself. It is high time to put an end to the cult of conventional probabilistic risk
modeling which is not able to cope with the kind of low-probability, high-impact
events that characterize systemic risk, in particular.
In this dissertation, our aim is threefold. Our first result is a negative one:
We argue – from a complexity and systems science perspective – that modern
financial systems are dynamically and organizationally complex which requires
explanatory models for an effective and successful approach to the management
of systemic risks. Yet, risk evaluations based on statistical calculations are not
sufficiently explanatory. Even worse for the mainstream, this study undermines
the citadel of risk assessment and management, arguing that probability theory is
not an adequate foundation for modeling systemic and extreme risk (in a banking
context).
Secondly, we present a kind of risk model on the constructive side which is
novel on two accounts. Firstly, in contrast to standard risk modeling approaches
used by banks, our proposal focuses on the knowledge dimension (banks’ assets
and liabilities are known) rather than the speculation dimension (unknown min-
imum or expected losses). Furthermore, in contrast to risk models aimed at regu-
lators who analyze systemic risk in order to preserve financial stability, our ob-
jective is not to model the financial system and its changes, but market
participants’ own positions and their propensity to react to outside changes. Our
plea for symbolic and logic-based risk modeling is illustrated with the example
of a certain credit crisis (Long-Term Capital Management) which we regard as a
systemic risk event in terms of its high rareness and its serious or extreme conse-