Kevin Dowd – An Introduction to Market Risk Measurement

DESCRIPTION

• Includes a CD-ROM that contains Excel workbooks and a Matlab manual and software.
• Covers the subject without advanced or exotic material.

TABLE OF CONTENTS

Preface xi

Acknowledgements

The risk measurement revolution.

The factors that contribute to it are 1.1.

A Volatile Environment

Growth in trading activity

There have been advances in information technology.

Risk measurement before VaR 3

Gap analysis 3

The analysis has a duration of 4.

Scenario analysis 4

Portfolio Theory 5.

Risk measures for derivatives

The value at risk is 7.

The origin and development of VaR 7

There are attractions of VaR 10.

There are criticisms of VaR 11.

The recommended reading is 12.

Measures of financial risk 13

The Mean–Variance Framework for Measuring Financial Risk 13 is a framework for measuring financial risk.

The Normality Assumption 13

2.1.2 Limitations of the Normality Assumption

Traditional Approaches to Financial Risk Measurement 18

Portfolio Theory 18

Approaches to Fixed-income Risk Measurement 18

The value at risk is 2.2.

VaR basics 19

VaR Parameters 24 have a choice.

Limitations of VaR as a Risk Measure 25

There is a VaR Uninformative of Tail Losses.

Perverse incentive structures can be created by VaR.

Diversification can be discouraged by VaR.

VaR Not Sub-additive 27

The expected tail loss is 28.

Coherent Risk Measures 28 are used.

The expected tail loss is 29

There are 33Conclusions.

The recommended reading is 33.

There are 3 basic issues in measuring market risk.

Data 35 is the data.

The profit/loss data is 35.

3.1.2 loss/profit data

3.1.3 returns data

Geometric returns data.

Estimating Historical Simulation VaR 36

Estimating Parametric VaR 37

Estimating VaR with NormallyDistributed Profits/Losses 38

Estimating VaR with Normally Distributed Arithmetic returns 39

Estimating Lognormal VaR 40

Estimating expected tail loss.

The summary was 3.5

There are positions to be mapped to risk factors.

Core instruments or factors can be selected.

Selecting Core Instruments 46

A3.1.2 Selecting Core Factors.

VaR Estimation 47 and a mapping positions.

The basic building blocks.

Basic foreign exchange positions 47

Basic Equity Positions 48

Zero-coupon bonds are 50.

Basic forward/futures 51

There are more complex positions.

A 3.3 recommended reading.

There are 4 non-parametric VaRs.

Compiling historical simulation data 55

Estimation of Historical Simulation VaR.

Basic Historical Simulation 56.

Estimating Curves and Surfaces for VaR.

Estimating Confidence Intervals for Historical Simulation VaR.

A quantile standard error approach to the estimation of confidence intervals is used.

There is an order statistics approach to the estimation of confidence intervals.

There is a bootstrap approach to the estimation of confidence intervals.

Weighted Historical Simulation 61.

The age-weighted historical simulation was 62.

Historical Simulation 63 is a volatility-weighted model.

The Historical Simulation 64 was Filtered.

Historical Simulation 66 has some advantages and disadvantages.

There are 66 advantages.

There are 4.5.2 Disadvantages.

The data set 67 has total dependence on it.

There are problems with the data period length.

The principal components are VaR and ETL.

Conclusions 69

Recommended reading is 70.

5 Parametric VaR.

Normal VaR and ETL 72.

General features 72

Disadvantages of Normality 76

The student t-distribution was 77.

The normal distribution is 78.

Extreme Value Distributions 81

The generalised extreme value distribution is 81.

The Peaks Over Threshold is a generalised Pareto approach.

The Multivariate NormalVariance– Covariance Approach 84 is used.

There are 86Conclusions

There is a recommended reading of 87.

Appendix: Delta–Gamma and Related Approximations.

A 5.1 Delta–normal Approaches.

A5.2 Delta–Gamma Approaches.

The Delta–Gamma Approximation is 90.

A5.2.2 The Delta–Gamma Normal Approach.

The Delta–Gamma Approach was written by Wilson.

Other Delta–Gamma Approaches 93

A5.3Conclusions

The recommended reading is 95.

There are 6 simulation approaches to VaR.

VaR and ETL 97 are options.

There are 97 preliminary considerations.

Estimating the VaR and ETL of an American Put 98 is an example.

There is an estimation of options VaR and ETL 99.

Estimating VaR is done using principal components.

Basic principal components simulation 99

There is a scenario simulation 100.

6.4 Fixed-income VaR.

The general considerations are as follows.

There are Stochastic processes for interest rates.

The term structure of interest rates.

There is a general approach to fixed-income VaR.

Estimating VaR and ETL is part of a Dynamic Portfolio Strategy 105.

Estimating Credit-related Risks with Simulation Methods

Estimating insurance risks using simulation methods.

Simulation methods are used for estimating pensions.

Estimating risks of defined-benefit pension plans

Estimating risks of defined-contribution pension plans

Conclusions 115

The recommended reading is 115.

There are 7 Incremental and Component risks.

VaR 117 is anIncremental VaR.

Incremental VaR 117 is interpreted.

Estimating IVaR is done by Brute Force.

Estimating I VaR using Marginal VaRs.

Garman’s approach is called thedelVaR.

The delVaR Approach 122 has potential Drawbacks.

Component VaR 122.

The properties of component VaR 122.

Component VaR 124 is used.

There is a drill-down capability.

Reporting Component VaRs 125

Conclusions 126

Recommended reading 126.

Estimating Liquidity Risks 127

There are risks to Liquidity and Liquidity Risks.

Estimating VaR and ETL 128

A Transactions Cost Approach 128

The approach is called the exogenous spread approach.

The market price response approach was used.

Derivatives Pricing Approaches 132

The discount approach is called the Liquidity Discount Approach.

There is a summary and comparison of alternative approaches.

The estimate of Liquidity at Risk is 135.

Estimating Liquidity in Crises 137

Recommended reading is 139.

Market risk models are backtested.

There are 141 preliminary data issues.

Obtaining Data 141

Statistical Backtests are based on the Frequency of Tail Losses.

The Basic Frequency-of-tail-losses is a test.

There is a Time-to-first-tail-loss Test.

A Tail-loss Confidence-Interval Test 146.

The approach to the backtesting is called the Christoffersen approach.

There are backtests based on the sizes of tail losses.

The basic sizes-of-tail-losses test is used.

The Crnkovic–Drachman Backtest Procedure was used.

The Berkowitz Approach is referred to as The Berkowitz Approach.

Evaluation Approaches to Backtesting is a forecast.

Basic ideas 153

The Frequency-of-tail-losses (Lopez I) Approach 154

The Size-adjusted Frequency (Lopez II) Approach 154

The Blanco–Ihle Approach 155 is referred to as The Blanco–Ihle Approach.

An alternative sizes-of-tail-losses approach was used.

There are other methods of comparing models.

Assessing the accuracy of backtest results

Testing with Alternative Confidence Levels, Positions and Data.

Backtesting with alternative confidence levels.

Backtesting with alternative positions.

Backtesting with alternative data.

Summary 159.

The recommended reading is 160.

10 stress tests

Benefits and Difficulties of Stress Testing

There are benefits to stress testing.

Difficulties with stress tests

Scenario analysis 167

Choosing Scenarios 167

Stylised scenarios 167

Actual Historical Events 168

There are 170 Hypothetical One-off Events.

The effects of scenarios are evaluated.

Testing for mechanical stress

Factor push analysis 172

The maximum loss-optimization number is 174.

Conclusions 175

The recommended reading is 175.

11 model risk

There are models and model risk.

The models are 11.1.1.

There is a model risk of 178.

There are sources of model risk.

The model specification is incorrect.

The model application is incorrect.

Implementation risk is 181

There are other sources of model risk.

Incorrect Calibration 182 was reported.

There are 182 programming problems.

There are 183 problems with the data.

Combating model risk is 183.

There are some guidelines for risk practitioners.

Guidelines for Managers for Combating Model Risk

There are methods to combat model risk.

Procedures to check and review models.

Independent Risk Oversight 187

There are 182Conclusions

Recommended reading is 188.

Toolkit 189.

There is a bibliography.

The author index is 271.

There is a subject index.

The software index is 283.

AUTHOR INFORMATION

Kevin didwd. He is a member of the Centre for Research in Risk and Insurance Studies at the University Business School.

THE WILEY ADVANTAGE

• Includes a CD-ROM that contains Excel workbooks and a Matlab manual and software.
• Covers the subject without advanced or exotic material.