Modelling Extremal Events

Contents

Reader Guideline

1. Risk Theory

The Ruin Problem

The Cramer-Lundberg Estimate

Ruin Theory for Heavy-Tailed Distributions

The Cramer-Lundberg for Large Claims

2. Fluctuations of Sums

The Laws of Large Numbers

The Central Limit Problem

Refinements of the CLT

The Functional CLT: Brownian Motion Appears

Random Sums

3. Fluctuations of Maxima

Limit Problems for Maxima

Weak Convergence of Maxima Under Affine Transformations

Maximum Domains of Attraction and and Norming Constants

Generalized Extreme Value Distribution and Generalized Pareto Distribution

Almost Sure Behavior of Maxima

4. Fluctuations of Upper Order Statistics

Order Statistics

The Limit Distribution of Upper Order Statistics

The Limit Distribution of Randomly Indexed Upper Order Statistics

Some Extreme Value Theory for Stationary Sequences

5. An Approach to Extremes via Point Processes

Basic Facts About Point Processes

Weak Convergence of Point Processes

Point Processes of Exceedances

Applications of Point Process Methods to IID Sequences

Some Extreme Value Theory for Linear Processes

6. Statistical Methods for Extremal Events

Exploratory Data Analysis for Extremes

Parameter Estimation for the Generalized Extreme Value Distribution

Estimating Under Maximum Domain of Attraction Conditions

Fitting Excesses Over a Threshold

7. Time Series Analysis for Heavy-Tailed Processes

A Short Introduction to Classical Time Series Analysis

Heavy-Tailed Time Series

Estimation of the Autocorrelation Function

Estimation of the Power Transfer Function

Parameter Estimation for ARMA Processes

Some Remarks about Heavy0Tailed Nonlinear Models

8. Special Topics

The Extremal Index

A Large Claim Index

When and How Ruin Occurs

Perpetuities and ARCH Processes

On the Longest Success Run

Some Results on Large Deviations

Reinsurance Treaties

Stable Processes

Self Similarity

Extreme value theory (EVT) is the study of probabilistic extremes. For example, if you were to randomly draw 10,000 standard normal variates, the maximum of those values is a random variable. EVA would answer questions related to the mean or standard deviation of that maximum value. EVA focuses primarily on asymptotic behavior as sample sizes approach infinity.

This is the authoritative text on EVT. Chapter 1 covers ruin theory as developed for insurance. The next three chapters discuss fluctuations of sums, maxima, and upper order statistics. Chapter 5 covers point processes. There are two chapters on statistical and time-series methods. Chapter 8 closes with special topics, such as:

 

when and how ruin occurs

ARCH processes,

the longest success run, and

reinsurance treaties and stable processes.

Appendices review topics from advanced probability theory.

This is the definitive text on EVT, but it may pose a challenge for readers who are new to the subject. An opening 20-page "reader guideline" may help. It gives an overview of the material before the start of the main text. You also might start with a more accessible book, such as Kotz and Nadarajah (2000), and then graduate to this one. As a reference, Embrechts et al. is unsurpassed.

For related books, see sections:

Math - Extreme Value Theory

Math - Probability

Math - Statistics

Math - Time Series Analysis

Risk Management - Market Risk

Risk Management - Operational Risk

 

 

Ads by Contingency Analysis.

Advertise on this site.

 

disclaimer

website: http://www.contingencyanalysis.com
books direct link: http://www.riskbook.com
copyright © Contingency Analysis, 1996 - current