12. Catastrophe, Climate and Emerging Risk Management

Introduction and Context

Insurance is designed to transform uncertainty into manageable financial consequences. Catastrophe and emerging risks challenge this function because they are often severe, correlated, rapidly changing, geographically concentrated, and difficult to estimate from historical experience. A flood may affect thousands of policyholders simultaneously; a pandemic may disrupt mortality, health, business interruption and financial markets at the same time; a cyber incident may spread across jurisdictions within minutes. Climate change further alters the frequency, severity and location of physical hazards while creating transition and liability risks.

For insurers and reinsurers, the central problem is not simply whether a peril exists. It is whether the peril can be identified, measured, priced, diversified, controlled and transferred without threatening solvency or undermining social usefulness. Catastrophe risk management therefore connects underwriting, accumulation control, capital management, reinsurance, claims preparedness, enterprise risk management and public policy.

The subject has contemporary importance because traditional assumptions are becoming less reliable. Historical loss data may not represent future climate conditions; supply-chain dependencies may create hidden concentrations; digital interconnectedness can turn a local cyber weakness into a systemic event; and protection gaps leave households, businesses and governments exposed even where insurance products are available.

Learning Objectives

After studying this chapter, the learner should be able to:

1. Explain the characteristics of catastrophe, climate and emerging risks.

2. Distinguish hazard, exposure, vulnerability, peril, accumulation and systemic risk.

3. Apply exposure management, catastrophe modelling and stress testing to insurance decisions.

4. Evaluate natural catastrophe, pandemic, cyber, terrorism, geopolitical and supply-chain risks.

5. Explain model uncertainty, parameter uncertainty, protection gaps and insurability constraints.

6. Assess the role of reinsurance, capital, governance and public-private cooperation.

7. Recommend practical and ethical risk-management responses for insurers and corporate risk managers.

The Risk Landscape: From Isolated Losses to Interconnected Events

A conventional insured loss is often treated as an independent event affecting a limited number of policies. Catastrophe risk is different. It is an event, or sequence of events, capable of producing unusually large aggregate losses. The defining feature is accumulation: many insured objects, contracts or portfolios are affected by one underlying cause.

A useful analytical distinction is between hazard, exposure and vulnerability. Hazard refers to the potentially damaging physical or human phenomenon, such as windstorm, flood, earthquake, epidemic disease or ransomware. Exposure refers to the people, property, infrastructure, economic activity or insured values located in the path of the hazard. Vulnerability describes the degree to which exposed assets are likely to be damaged or disrupted. Risk emerges from the interaction of these elements, together with financial terms such as limits, deductibles, exclusions and policy conditions.

Catastrophe risk is not limited to natural events. Terrorism, political violence, cyber attacks, pandemics and failures of critical infrastructure can produce similar accumulation effects. Emerging risk refers to a newly developing or changing risk whose implications are not yet sufficiently understood, measured or priced. It may be a genuinely new threat, such as a novel technology, or a familiar risk whose scale and transmission mechanisms have changed.

Several characteristics make these risks difficult to manage:

• Low frequency and high severity: reliable statistical estimation is difficult when observations are scarce.
• Correlation: losses occur across many policies or lines of business at the same time.
• Non-stationarity: the relationship between past and future events may change.
• Complex causation: losses may result from interacting physical, technological, economic and social factors.
• Rapid transmission: digital and supply-chain networks spread disruption beyond the original location.
• Ambiguous policy wording: coverage may depend on causation, exclusions, aggregation and proof of loss.
• Social and political sensitivity: decisions concerning availability, affordability and claims settlement attract public scrutiny.

A Taxonomy of Contemporary Catastrophe and Emerging Risks

Natural catastrophe risk

Natural catastrophe risk includes earthquakes, tropical cyclones, floods, severe convective storms, wildfires, landslides, volcanic eruptions, tsunamis and other hazards. The insurer must consider not only the primary peril but also secondary perils and cascading effects. A cyclone may create storm surge, inland flooding, wind damage, business interruption and public infrastructure failure. An earthquake may trigger fire, liquefaction, landslides and prolonged loss of access.

Climate risk

Climate risk is commonly divided into physical, transition and liability risk. Physical risk arises from acute events such as storms and floods or chronic changes such as sea-level rise, heat and water stress. Transition risk results from changes in law, technology, consumer preference and asset values during movement toward a lower-carbon economy. Liability risk arises when parties face claims for contributing to climate-related harm or failing to disclose material risks.

Climate risk affects underwriting, reserving, investments, reinsurance costs, asset-liability management and solvency. It may also create a feedback loop: repeated losses reduce insurance availability, which lowers property values and weakens investment in resilience, thereby increasing future vulnerability.

Pandemic and biological risk

Pandemics combine health, economic and operational risks. They may generate mortality and morbidity claims, medical expenses, event cancellation losses, travel claims, business interruption disputes, employee absence and investment volatility. Unlike many natural catastrophes, a pandemic can last for months or years and may affect insurers, reinsurers, service providers, courts and regulators simultaneously. Policy triggers and exclusions are therefore especially important.

Cyber risk

Cyber risk includes confidentiality breaches, data theft, system interruption, extortion, fraud, technology failure and damage caused through digital dependency. Cyber accumulation is difficult because many insureds may rely on the same cloud provider, software platform, telecommunications network or data centre. A vulnerability in one widely used product can produce correlated losses across industries and countries.

Cyber risk also illustrates the difference between silent and affirmative exposure. Affirmative cyber coverage is expressly provided under a policy. Silent cyber exposure arises when traditional policies, such as property, liability or marine contracts, unintentionally respond to cyber-related loss. Managing silent cyber requires wording reviews, scenario analysis and portfolio-wide identification of digital dependencies.

Terrorism and political violence

Terrorism risk is characterized by intentional action, uncertain frequency, symbolic targets and potential concentration in urban, transport, energy or financial locations. Political violence, civil disorder, war and state-sponsored cyber operations may overlap but require careful legal and underwriting distinctions. Accumulation controls must consider location, target type, event definition, hours clauses, aggregation and government compensation arrangements.

Geopolitical and sovereign risk

Geopolitical risks include sanctions, trade restrictions, armed conflict, expropriation, currency controls, political instability and abrupt regulatory change. These risks can affect credit, political risk, marine, aviation, energy, construction and investment portfolios. They are particularly challenging because political decisions can change rapidly and because losses may be correlated across countries or sectors.

Supply-chain and infrastructure risk

Modern production depends on interconnected suppliers, ports, logistics systems, energy grids, payment networks and digital platforms. A single bottleneck can interrupt many firms that are geographically dispersed. Insurance analysis must therefore examine contingent business interruption, supplier concentration, customer concentration, replacement lead times, alternative sourcing and critical infrastructure dependencies.

Exposure Management and Accumulation Control

Exposure management begins with a reliable inventory of what is insured, where it is located, how much it is worth and what contractual terms apply. Data should include geocoded addresses, construction type, occupancy, replacement value, policy limits, deductibles, business interruption values, location of suppliers and relevant hazard characteristics.

The sum insured is not necessarily the same as the insurer's realistic loss. A portfolio may contain deductibles, sublimits, exclusions, coinsurance, waiting periods and policy conditions. Exposure management therefore distinguishes between gross exposure, net exposure after contractual features, and probable financial loss after reinsurance.

Concentration can be measured by geography, peril, line of business, industry, counterparty, technology provider or event type. A portfolio may appear diversified by policy count but remain concentrated by insured value. For example, thousands of small commercial policies may depend on one electricity grid or cloud provider. Accumulation control must therefore examine common causes rather than only physical proximity.

Important management techniques include:

1. Geographic limits: restricting aggregate values in high-hazard zones.

2. Per-risk and per-event limits: controlling individual and aggregate exposures.

3. Deductibles and sublimits: retaining predictable losses while limiting severity.

4. Underwriting standards: requiring resilience features such as flood barriers, fire protection or cyber controls.

5. Portfolio diversification: balancing locations, industries, products and perils.

6. Reinsurance protections: transferring losses above specified attachment points.

7. Real-time monitoring: updating exposures after acquisitions, inflation, construction or portfolio growth.

Exposure data quality is a governance issue, not merely an administrative matter. Incorrect coordinates, outdated values and inconsistent classifications can materially distort pricing, capital and reinsurance decisions.

Catastrophe Models: Value and Limitations

A catastrophe model generally combines four components. The hazard module represents the frequency, intensity and geographic distribution of events. The exposure module describes insured assets and their characteristics. The vulnerability module estimates damage as a function of hazard intensity and asset features. The financial module applies deductibles, limits, policy conditions, insurance terms and reinsurance structures to convert damage into insured loss.

Models produce metrics such as occurrence exceedance probability, aggregate exceedance probability, probable maximum loss and annual average loss. An exceedance probability curve shows the likelihood that loss will exceed a particular amount. A one-in-two-hundred-year loss is commonly interpreted as an annual exceedance probability of approximately 0.5 percent, not as an event that occurs exactly once every two hundred years.

Models assist with pricing, underwriting, capital allocation, reinsurance purchase, portfolio steering and regulatory reporting. They are decision-support tools rather than substitutes for judgement. Model results may differ because of assumptions about event frequency, vulnerability, demand surge, inflation, secondary perils, climate change or policy response.

Three forms of uncertainty require attention. Aleatory uncertainty reflects natural randomness in events. Parameter uncertainty arises because estimated frequencies, severities or vulnerability relationships are imperfect. Model uncertainty arises because different models represent the hazard or financial mechanism differently. A prudent insurer uses sensitivity testing, alternative models, expert judgement and transparent documentation rather than treating one model output as a fact.

Stress Testing and Scenario Analysis

Stress testing asks how the organization would perform under severe but plausible conditions. It is especially useful when historical data are inadequate or when a risk is changing rapidly. A scenario should identify the initiating event, affected exposures, operational consequences, financial transmission channels, management actions and secondary effects.

A natural catastrophe scenario may combine wind damage, storm surge, claims inflation, reinsurance exhaustion and investment-market disruption. A cyber scenario may examine simultaneous interruption at a major technology provider, data-restoration costs, liability claims, regulatory penalties and reputational effects. A pandemic scenario may include prolonged claims development, staff absence, premium collection problems, litigation and reduced asset values.

Good scenarios are not forecasts. They are structured tests of resilience. They should include reverse stress testing, which asks what combination of events could threaten solvency, liquidity, operational continuity or regulatory compliance. Senior management should agree in advance on escalation thresholds and recovery actions.

Stress testing is most valuable when it changes decisions. Results may lead to revised underwriting limits, additional reinsurance, capital buffers, liquidity arrangements, supplier diversification, wording changes or investment adjustments. A scenario that merely produces a report without management response has limited risk-management value.

Insurability, Protection Gaps and the Role of Reinsurance

A risk is more readily insurable when losses are definable, accidental, measurable, financially significant but not ruinous, and sufficiently diversified. The law of large numbers is more effective when risks are numerous and reasonably independent. Catastrophe and emerging risks often violate these conditions because they are correlated, uncertain and potentially systemic.

Insurability may be constrained by:

• lack of credible data;
• rapid changes in hazard or technology;
• extreme accumulation;
• moral hazard and adverse selection;
• difficulty proving causation;
• legal uncertainty;
• unaffordable premiums;
• inability to model policy exclusions consistently; and
• potential exhaustion of available capital.

A protection gap is the difference between economic losses and insured losses. It may arise from lack of awareness, affordability problems, exclusions, underinsurance, informal development, limited distribution or absence of suitable products. Protection gaps have social consequences because uninsured losses can reduce household wealth, delay recovery and increase pressure on governments.

Reinsurance contributes through risk transfer, capital relief, capacity creation and technical expertise. Proportional treaties may support portfolio growth and sharing of premiums and losses. Non-proportional catastrophe excess-of-loss cover protects against event losses above an attachment point. Aggregate covers respond when total losses during a period exceed an agreed threshold. Parametric reinsurance pays on the occurrence of an objective index, such as wind speed or earthquake intensity, rather than on adjusted physical loss. It can provide rapid liquidity but may create basis risk when the index does not closely match the insured's actual loss.

Alternative capital, including catastrophe bonds and collateralized reinsurance, can add capacity by linking insurance risk with capital-market investors. However, basis risk, collateral structure, trigger design, investor appetite and transaction complexity require careful evaluation.

Governance, Regulation and Professional Judgement

Catastrophe and emerging risk management requires board oversight because extreme events can threaten strategy, solvency and public trust. Governance should define risk appetite, concentration limits, escalation procedures, model-use standards, data responsibilities and reinsurance objectives.

The three-lines model is useful. Operational and underwriting functions own and manage risk; independent risk and compliance functions challenge assumptions and monitor limits; internal audit provides assurance over governance and controls. Actuarial, catastrophe-modelling, claims, legal, technology, investment and business-continuity specialists should work together rather than operating in isolated departments.

Regulatory expectations commonly emphasize solvency, own-risk-and-solvency assessment, operational resilience, outsourcing controls, cyber security, climate disclosures, fair treatment of customers and credible claims-paying capacity. Exact requirements differ by jurisdiction, but the professional principle is consistent: material assumptions must be documented, challenged and communicated.

Ethical issues are significant. Risk-based pricing can improve technical adequacy but may make essential insurance inaccessible to vulnerable communities. Automated underwriting may reproduce historical discrimination if data reflect unequal investment or biased claims experience. Claims automation must allow review, explanation and correction. Insurers should not exaggerate certainty in catastrophe models or use exclusions that are technically valid but incomprehensible to policyholders.

Digital Tools, Analytics and Artificial Intelligence

Geospatial information systems support hazard mapping, exposure visualization and accumulation monitoring. Remote sensing and satellite imagery can improve property characteristics and post-event claims assessment. Application programming interfaces can link underwriting systems with weather, geocoding, infrastructure and security data. Machine learning may identify patterns in claims, cyber signals or supply-chain relationships.

These tools do not eliminate uncertainty. Data may be incomplete, biased or unstable after a major event. Artificial intelligence can produce opaque decisions, false correlations and privacy risks. Model governance should therefore include validation, performance monitoring, explainability appropriate to the use case, human oversight, access controls and an audit trail. AI should augment professional judgement, not conceal accountability.

Illustrative Case: Coastal Insurer Facing Compound Catastrophe Risk

Background

HarbourShield, a fictional regional insurer, writes residential and small-business property insurance along a rapidly developing coastline. Its portfolio has grown quickly through digital distribution. A catastrophe model indicates substantial windstorm exposure, but internal data show that many properties also face storm surge and inland flooding. Several commercial policyholders depend on the same port and electricity network.

Situation and problem

Management initially focuses on the modelled windstorm loss. A review identifies three hidden concentrations: high insured values in a narrow coastal corridor, contingent business interruption linked to the port, and common dependence on one cloud-based claims platform. Climate scenarios also suggest that historical flood experience may understate future losses. Reinsurance purchased on a wind-only basis may not respond adequately to a compound event.

Analysis

The risk function constructs an exposure database using geocoded locations, updated replacement costs, occupancy information and supplier dependencies. It runs scenarios combining wind, surge, inland flood, port closure, claims inflation and a temporary outage of the claims platform. Alternative hazard assumptions are tested because the model does not represent future climate conditions with precision.

The analysis shows that the principal threat is not one isolated peril but correlated accumulation. Gross loss could be large, while liquidity pressure could arise before all claims are settled. A wind-only treaty leaves basis and coverage gaps. The business interruption portfolio also contains ambiguous wording concerning damage to external infrastructure.

Approach and strategy

HarbourShield adopts a multi-part response:

• tighter coastal accumulation limits and revised underwriting authority;
• property-level flood questions and resilience requirements;
• updated values and supplier-dependency data;
• a multi-peril catastrophe excess-of-loss programme;
• a separate aggregate protection layer for repeated secondary events;
• a contingency arrangement for claims processing if the cloud platform fails;
• clearer business interruption wording and customer communication;
• quarterly board reporting on climate scenarios and protection gaps.

The insurer retains ordinary attritional losses but transfers severe event risk above a defined attachment point. It also works with local authorities and lenders on mitigation, recognizing that insurance alone cannot reduce physical vulnerability.

Outcome

The programme increases the cost of reinsurance and requires some portfolio reduction, but it improves the relationship between exposure, capital and risk appetite. Management gains a clearer view of liquidity needs and silent dependencies. Some customers face higher deductibles, while resilience improvements enable others to obtain more stable coverage.

Learning points

The case demonstrates that accumulation is multidimensional. Geographic concentration, common infrastructure, digital dependence and policy wording can interact. It also illustrates why model outputs should be supplemented by scenario analysis, data improvement, alternative assumptions and managerial judgement.

Common Mistakes and Practical Best Practices

A frequent mistake is relying on historical averages without asking whether the underlying risk is changing. Another is treating modelled probable maximum loss as a precise prediction. Insurers also underestimate secondary perils, inflation, demand surge, non-damage business interruption and operational disruption during claims settlement.

Other weaknesses include incomplete geocoding, failure to update sums insured, excessive dependence on one reinsurer or vendor, untested exclusions, inadequate liquidity planning and poor communication between underwriting and risk functions. Reinsurance may be purchased because it is customary rather than because it fits the actual loss distribution and risk appetite.

Strong practice includes:

1. Maintain a living exposure register with data ownership and quality controls.

2. Monitor concentrations by cause, not merely by location or policy count.

3. Compare multiple models and document material assumptions.

4. Use severe-but-plausible and reverse stress scenarios.

5. Integrate underwriting, claims, investments, operations and reinsurance decisions.

6. Test wording against realistic event sequences and legal interpretations.

7. Plan liquidity, communications, outsourced services and claims continuity.

8. Review protection gaps and affordability alongside profitability.

9. Govern AI and external data with validation and human accountability.

10. Treat risk mitigation and adaptation as complements to insurance transfer.

Current Developments and Emerging Directions

Climate-adjusted underwriting, resilience-linked pricing and public-private catastrophe pools are receiving increasing attention. Parametric products are being developed for rapid post-event liquidity, particularly where traditional loss adjustment is slow or data are limited. Cyber underwriting is becoming more disciplined as insurers seek clearer aggregation controls and security requirements.

Scenario-based climate analysis is increasingly used alongside traditional financial reporting. Insurers are also examining nature-related risks, biodiversity loss, water stress, misinformation, autonomous systems, space infrastructure and the security of critical digital services. These areas remain uncertain, but early governance is preferable to waiting until exposures become material and difficult to control.

The broad direction of professional practice is toward integrated resilience. The insurer of the future must understand physical assets, financial contracts, digital networks, social vulnerability and public policy as parts of one risk system.

Chapter Summary

Catastrophe, climate and emerging risks challenge insurance because they are severe, correlated, dynamic and uncertain. Effective management begins with distinguishing hazard, exposure and vulnerability, then building accurate exposure data and controlling accumulation by geography, peril, industry, infrastructure and technology dependency. Catastrophe models support pricing, capital and reinsurance decisions, but their outputs must be interpreted through model, parameter and aleatory uncertainty.

Stress testing and scenario analysis are essential where historical experience is inadequate, especially for climate change, pandemics, cyber events and geopolitical disruption. Reinsurance, alternative capital and parametric structures can transfer or finance severe losses, although basis risk and capacity constraints remain important. Protection gaps and insurability limitations require cooperation among insurers, reinsurers, governments, communities and businesses. Strong governance combines board oversight, data discipline, professional challenge, ethical treatment of customers, operational resilience and responsible use of technology.

Key Terms

Hazard: A potentially damaging physical, biological, technological or human phenomenon.

Exposure: People, property, activities or values that may be affected by a hazard.

Vulnerability: The susceptibility of an exposed asset or system to damage or disruption.

Catastrophe risk: Risk of an event or sequence of events producing unusually large aggregate loss.

Accumulation: Concentration of insured values or liabilities exposed to a common cause of loss.

Emerging risk: A developing or changing risk whose effects, probability or measurement are not yet sufficiently understood.

Physical climate risk: Loss arising from acute or chronic physical effects of climate conditions.

Transition risk: Financial risk arising from economic, technological, legal or policy changes during transition to a lower-carbon economy.

Protection gap: The difference between total economic loss and the portion covered by insurance.

Catastrophe model: A model combining hazard, exposure, vulnerability and financial components to estimate event losses.

Probable maximum loss: An estimate of loss from a severe event at a specified level of probability or confidence.

Exceedance probability: The likelihood that loss will exceed a specified amount during a defined period.

Model uncertainty: Uncertainty arising because alternative models represent risk differently.

Basis risk: The possibility that an index or trigger payment does not match actual insured loss.

Silent cyber: Unintended cyber exposure within policies not expressly designed to cover cyber risk.

Stress testing: Examination of financial or operational performance under severe but plausible conditions.

Parametric insurance: Cover that pays when a specified measurable index or event threshold is reached.

Systemic risk: Risk that disruption spreads through interconnected institutions, networks or markets.

Reverse stress testing: Analysis that identifies conditions capable of causing failure or breaching a critical limit.

Contingent business interruption: Loss caused by disruption at a supplier, customer or external infrastructure provider.

Review Questions

Short Answer Questions

1. Distinguish hazard, exposure and vulnerability.

2. What is meant by accumulation in catastrophe insurance?

3. Explain the difference between physical and transition climate risk.

4. Define silent cyber exposure.

5. What is basis risk in parametric insurance?

Long Answer Questions

1. Explain the structure, uses and limitations of catastrophe models.

2. Discuss how insurers can manage natural catastrophe and climate accumulation.

3. Examine the factors that make pandemic and cyber risks difficult to insure.

4. Explain the role of reinsurance and alternative capital in catastrophe risk management.

5. Discuss governance and ethical issues in the use of catastrophe analytics and artificial intelligence.

Analytical / Application Questions

1. An insurer has a geographically diversified portfolio, but most policyholders depend on one cloud provider and one electricity grid. Analyse the hidden accumulation and recommend controls.

2. A coastal property portfolio shows low historical flood losses but high projected future exposure. Design a stress-testing and underwriting response that addresses model uncertainty.

3. A government asks insurers to close a growing disaster protection gap without making cover unaffordable. Evaluate a possible public-private insurance and resilience framework.

22. Multiple Choice Questions with Answers

Question 1

Which combination best explains catastrophe accumulation?

A. Many independent claims with stable severity

B. Many exposures affected by a common event or cause

C. Only claims arising from fraud

D. Losses restricted to one individual policy

Correct Answer: B. Many exposures affected by a common event or cause

Question 2

Which statement best describes transition climate risk?

A. Damage caused directly by a cyclone

B. Loss caused by earthquake aftershocks

C. Financial effects of policy, technology or market changes during decarbonisation

D. Claims arising from ordinary wear and tear

Correct Answer: C. Financial effects of policy, technology or market changes during decarbonisation

Question 3

What is the main purpose of the vulnerability module in a catastrophe model?

A. To identify the insurer's shareholders

B. To estimate damage given hazard intensity and asset characteristics

C. To determine the legal jurisdiction of a policy

D. To calculate investment income

Correct Answer: B. To estimate damage given hazard intensity and asset characteristics

Question 4

Why can a portfolio with geographically dispersed insureds still have cyber accumulation?

A. Cyber losses are always limited to one building

B. Insureds may depend on the same software, cloud or telecommunications provider

C. Cyber policies cannot contain exclusions

D. Geographic diversification eliminates digital dependency

Correct Answer: B. Insureds may depend on the same software, cloud or telecommunications provider

Question 5

A one-in-two-hundred-year loss estimate generally represents approximately:

A. A loss guaranteed to occur once every 200 years

B. An annual exceedance probability of about 0.5 percent

C. A 200 percent annual probability

D. The average annual loss over 200 years

Correct Answer: B. An annual exceedance probability of about 0.5 percent

Question 6

Which feature most clearly distinguishes stress testing from forecasting?

A. Stress testing examines severe plausible conditions rather than predicting the most likely outcome

B. Stress testing never uses assumptions

C. Forecasting is always more uncertain than stress testing

D. Stress testing applies only to investment portfolios

Correct Answer: A. Stress testing examines severe plausible conditions rather than predicting the most likely outcome

Question 7

What is basis risk in a parametric catastrophe cover?

A. The reinsurer's failure to collect premium

B. A mismatch between the index-triggered payment and the insured's actual loss

C. The risk that a policy has no deductible

D. The risk that a model uses too many historical observations

Correct Answer: B. A mismatch between the index-triggered payment and the insured's actual loss

Question 8

Which action best addresses silent cyber risk?

A. Ignoring cyber exposure in traditional policies

B. Reviewing wording and identifying unintended cyber exposure across all lines

C. Removing all technology from underwriting systems

D. Using only historical fire losses for pricing

Correct Answer: B. Reviewing wording and identifying unintended cyber exposure across all lines

Question 9

Why are protection gaps a public-policy concern?

A. They ensure that all economic losses are insured

B. They can delay recovery and transfer financial pressure to households and governments

C. They eliminate the need for mitigation

D. They occur only in highly insured economies

Correct Answer: B. They can delay recovery and transfer financial pressure to households and governments

Question 10

Which is the strongest management response to model uncertainty?

A. Treating one model output as exact

B. Using sensitivity analysis, alternative models and documented expert judgement

C. Avoiding all catastrophe insurance

D. Replacing exposure data with premium volume

Correct Answer: B. Using sensitivity analysis, alternative models and documented expert judgement