Skip to content

GRA · Actuarial and Quantitative Risk

Spatial financeWhere capital meets hazard, exposure and transition

Extreme value methods

Extreme value theory gives a statistical basis for estimating the size and frequency of losses and hazards beyond what the record shows. It is used for large-loss pricing, catastrophe hazard, operational risk and capital at high confidence levels. Results depend on thresholds, sample size and whether conditions are stable. This page sets out how extreme value methods work, where they fail, the data they need and who relies on them.

01Canvas

How the work runs.

Systems · around this subjectIllustrative
Reserving workbench: the screen’s structure, terms and states for Extreme value methods.Example names and illustrative values; a design, not a record or a live system.

Fig.Reserving workbench: example names and illustrative values, not a record or a live system.

02The question

Extreme value methods: how the risk develops.

Tail estimation for hazards, markets and operational losses. The dossier below sets out how the risk develops, what is measured, where the data come from, how it is modelled and who decides.

Mechanism

Tails follow limiting laws

Under broad conditions, block maxima follow a generalised extreme value distribution and exceedances over a high threshold follow a generalised Pareto distribution.

Mechanism

Thresholds trade bias for variance

A high threshold respects the theory but leaves few points; a low threshold adds data but bias. Results shift with the choice.

Mechanism

Short records widen intervals

Estimating a rare return period from a short record gives wide confidence intervals that are often left unreported.

Mechanism

Non-stationarity breaks assumptions

If climate, exposure or claims practice change, past extremes may not represent future ones, and stationary fits mislead.

03Exposure, loss and capital

What extreme value methods means for exposure, loss and capital.

The exposure, loss, scenario and capital views a risk officer, supervisor or ministry of finance would ask for, with the frameworks they report against. Structures follow supervisory and actuarial practice. Values are illustrative: never prices, ratings or advice.

Exposure and loss at return periods

How losses from extreme value methods grow with severity, gross and net of reinsurance, against capital.

1 in 101 in 501 in 1001 in 250grossnet of reinsurance800
Return periodAnnual probabilityGross lossNet lossNet, capital = 100
1 in 100.10141010
1 in 500.02332121
1 in 1000.01462727
1 in 2500.004683939
Average annual loss3.12.22.2
Illustrative loss index for a example portfolio (capital = 100): the exceedance-probability form insurers and supervisors read. Not data on any institution, not a price or a rating.
NGFS climate scenarios · Network for Greening the Financial SystemIFRS S2 Climate-related Disclosures · ISSBTCFD recommendations · Financial Stability Board TCFDPrinciples for the effective management of climate-related financial risks · Basel Committee on Banking Supervision
04Viashiria

What is measured, and how well.

Each indicator carries its unit, its cadence, the kind of source behind it and the evidence label a value would carry. No values are shown: a programme supplies them under its own authority.

IndicatorUnitCadenceSource typeLabel
Shape parameter of the taildimensionlessper analysisFitted extreme value modelsIMEUNDWA KIMUUNDO
Return level at stated return periodcurrency or hazard unitper analysisFitted modelsIMEUNDWA KIMUUNDO
Confidence interval width for return levelcurrency or hazard unitper analysisBootstrap or profile likelihoodIMEUNDWA KIMUUNDO
Exceedances above thresholdcountper analysisDataIMEONEKANA
Record lengthyearsper datasetDataIMEONEKANA
Trend in annual maximachange per decadeper analysisNon-stationary fitsIMEKADIRIWA
05Vyanzo

Where the evidence comes from.

Named public datasets and institutional records, each with its licence class. Sources by agreement are used only under their terms and never republished.

ChanzoWhat it givesLicence class
Coles, An Introduction to Statistical Modeling of Extreme ValuesStandard reference text on extreme value methodsProprietary, published book
R extRemes and evd packagesOpen-source extreme value toolsOpen source, GPL
NOAA and national meteorological hazard recordsLong records of hazard extremesPublic domain or terms per service
Operational risk loss data consortia publicationsLarge operational loss recordsMembers only; summaries free to read
Large-loss databases of insurers and reinsurersIndividual large claimsProprietary, internal
06Decisions

Who decides, and what the evidence serves.

Actuarial and Quantitative Risk prepares evidence. The decisions below are taken by the authorities named, under their own rules, and never by the platform.

Nani anaamuaThe questionEvidence this page serves
Actuaries and risk modellersWhich extreme value method and threshold to useStability diagnostics and data review
Capital model ownersHow tail estimates enter capitalUncertainty and sensitivity evidence
ValidatorsWhether tail fits are adequateBack-testing and alternative fits
SupervisorsWhether tail calibration supports regulatory capitalModel documentation and validation

Outputs this topic produces

  • Guide to block maxima and peaks-over-threshold methods
  • Method note on threshold selection and stability plots
  • Brief on reporting uncertainty for return levels
  • Questions for reviewers on non-stationarity
07Core questions

The questions experts ask.

  • How large is the loss that is exceeded, on average, once in a given number of years?
  • Which threshold or block size gives a reliable estimate of the tail?
  • Is the tail heavy, light or bounded, and how confident can we be?
  • Is the extreme behaviour changing over time?
  • How do extremes in different variables or places occur together?
08Mechanisms

The theory of extremes.

Two classical results underpin the field. The distribution of block maxima, such as the largest annual flood, tends towards the generalised extreme value distribution under broad conditions. The distribution of exceedances over a high threshold tends towards the generalised Pareto distribution. A single shape parameter determines whether the tail is heavy, light or bounded, and it is the most important and most uncertain quantity in any extreme value analysis.

Practical choices drive results: the block size or threshold, how to handle clusters of extremes that occur together, such as aftershocks or volatile trading days, and whether to include covariates for trends or seasonality. Non-stationary extreme value models let parameters change with time or with climate indices, which is essential when the underlying process is changing.

Multivariate and spatial extremes address joint behaviour, such as storm surge with rainfall, or losses across regions. Dependence in the extremes can differ from dependence in the bulk of the data, which has direct consequences for diversification.

09Methods and data

Methods and their limits.

Peaks

Peaks over threshold

Fitting the generalised Pareto distribution to exceedances; results depend on threshold choice.

Maxima

Block maxima

Fitting the generalised extreme value distribution to block maxima; uses data less efficiently.

Non-stationary

Covariate models

Letting parameters vary with time or climate; risks over-fitting short records.

Limits: estimates far beyond the data rest on the assumption that the limiting distribution applies, and confidence intervals for long return periods are wide. Reporting return levels without intervals overstates certainty.

11Maswali

Questions and answers.

What is the peaks-over-threshold method?
An extreme value approach that models losses above a high threshold with a generalised Pareto distribution, using more data than annual maxima.
What does the shape parameter tell us?
Whether the tail is heavy, light or bounded; a positive shape implies a heavy tail where very large losses are relatively more likely.
Can extreme value theory predict the next catastrophe?
No. It estimates frequencies and sizes of extremes under assumptions, with wide uncertainty, not when an event will happen.
What is a return period?
The average interval between exceedances of a given level under stationary conditions. It is not a schedule: a rare event can happen in consecutive years.
Can extreme value theory predict a crash?
No. It estimates the probability and size of extremes, not their timing.
What if the climate is changing?
Non-stationary models allow parameters to change, but they need care to avoid over-fitting short records.
RipotiChangia

Changia ushahidi kutoka ukurasa huu.

Reports ni chapisho la kila robo mwaka la Muungano. Unachojua kuhusu somo hili kinaweza kuwa ushahidi, kisa au sahihisho katika toleo lijalo, kikitajwa kwa jina lako na kuchapishwa chini ya CC BY 4.0 pale ambapo ni kazi yako mwenyewe.

Global Risks AllianceActuarial and Quantitative RiskMethods and TechnicalFinance and Insurance Readiness
Hatari za Kiaktuari na Kiidadi

Bring an exposure or capital question to the Alliance.

Information and enquiries are open today. An enquiry receives a reference and names the next step; applications, payments, compute allocations and registrations open only when their team, terms and review capacity exist.