3. Risk Pooling and Theoretical Foundations

Introduction and Context

Insurance is often described as a mechanism for transferring risk, but transfer alone does not explain how an insurer can accept risks that would be financially unbearable for an individual. The economic foundation of insurance is the pooling of many uncertain exposures. By combining risks that are sufficiently numerous and reasonably independent, an insurer converts a collection of uncertain individual outcomes into a more predictable aggregate result. This predictability supports pricing, reserving, underwriting, claims management and the continuing payment of compensation to policyholders.

Risk pooling is therefore both a mathematical and an institutional idea. Mathematically, it relies on probability, diversification and the law of large numbers. Institutionally, it depends on mutuality: policyholders contribute to a common fund from which the losses of the relatively few unfortunate members are paid. Insurance markets also require risk transfer, because the insurer assumes contractual responsibility for specified losses. Yet pooling is not automatic. Adverse selection, moral hazard, correlated losses, inaccurate information and inadequate governance can weaken the pool and threaten solvency.

This chapter develops the principal theoretical foundations of insurance portfolios. It explains how mutuality and risk transfer interact, why diversification improves predictability, how the law of large numbers supports actuarial pricing, and why information asymmetry creates adverse selection and moral hazard. It also considers the managerial, regulatory and technological implications of these ideas.

Learning Objectives

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

1. Explain the economic meaning of risk pooling, mutuality and risk transfer.

2. Distinguish between independent, dependent, systematic and idiosyncratic risks.

3. Apply diversification and the law of large numbers to insurance portfolios.

4. Analyse the effects of adverse selection and moral hazard on insurance markets.

5. Explain the underwriting, pricing, claims and governance responses used to control these problems.

6. Evaluate the limitations of pooling in the presence of catastrophe, common-cause and emerging risks.

7. Apply the theoretical concepts to managerial and professional insurance decisions.

The Architecture of an Insurance Pool

Risk, uncertainty and insurable exposure

Risk refers to the possibility of an undesirable outcome where the range of possible outcomes can be identified, at least approximately. In insurance, the relevant outcome is commonly a financial loss caused by events such as death, illness, accident, fire, theft, liability or natural catastrophe. A risk becomes an insurable exposure when it can be defined, measured and associated with a sufficiently credible loss experience, and when the proposed coverage is legally and commercially acceptable.

Insurance generally deals most effectively with pure risks, where the outcome is either a loss or no loss. Speculative risks, such as entrepreneurial investment, involve both gain and loss and are not ordinarily insured in their complete form. Even within pure risks, the insurer must establish whether the event is accidental, the loss is measurable, the exposure is sufficiently homogeneous for comparison and the premium can be determined.

Mutuality and the common fund

Mutuality is the principle that members of a group collectively bear the financial consequences of losses suffered by a minority of members. Each policyholder contributes a premium, but only some policyholders make claims during a particular period. The arrangement is not a promise that every contributor will receive an equivalent monetary return. Its value lies in replacing the possibility of a large, uncertain personal loss with a smaller, certain premium.

In a traditional mutual insurer, policyholders may have a direct ownership or participation interest in the insurance organisation. In a stock insurer, shareholders provide capital and own the company, while policyholders participate through contracts rather than ownership. Despite this legal distinction, both forms depend economically on pooling: premiums from a broad group support the payment of covered claims.

Mutuality also has a social dimension. Health insurance, agricultural insurance and life insurance may redistribute resources across different states of need, ages or risk experiences. Such redistribution is legitimate only when it is consistent with the contract, applicable regulation and transparent pricing arrangements. If cross-subsidies are hidden or excessive, lower-risk participants may withdraw, weakening the pool.

Risk transfer

Risk transfer occurs when one party shifts the financial consequences of a specified uncertain event to another party under a contract. In an insurance contract, the policyholder pays a premium and the insurer assumes responsibility for covered losses, subject to terms, conditions, exclusions, deductibles, limits and policy duration. The policyholder does not transfer the physical possibility of the event; a fire can still occur. What is transferred is the financial burden specified in the contract.

Risk transfer is incomplete in several important ways. Deductibles retain part of the loss with the insured. Policy limits cap the insurer’s responsibility. Exclusions remove particular causes or circumstances. Coinsurance and participation clauses may allocate a portion of the loss to the policyholder. These features are not merely contractual technicalities; they influence incentives, claims frequency and the quality of the pool.

Risk pooling and risk transfer are complementary. Transfer without pooling would leave the insurer exposed to the full volatility of each individual loss. Pooling without a contractual transfer mechanism would lack a clear allocation of responsibility. Insurance combines both: the policyholder transfers defined financial risk, while the insurer pools many transferred exposures.

Diversification, Independence and the Law of Large Numbers

Why many risks are more predictable than one risk

Consider an individual household facing a 1 per cent probability of a covered loss of 100,000 monetary units during a year. The expected loss is 1,000 monetary units, but the actual outcome is either zero or 100,000. For that household, the result is highly uncertain. If an insurer covers 10,000 similar and independent households, the expected aggregate loss is 10 million monetary units. The number of claims will still vary, but the proportion of households suffering a loss will tend to be closer to 1 per cent than it would be in a very small group.

If individual losses are independent, the variance of the aggregate loss is the sum of individual variances. The standard deviation of the aggregate tends to increase approximately with the square root of the number of risks, while the expected aggregate loss increases proportionately with the number of risks. Consequently, the relative volatility of the average loss declines as the portfolio grows. This is the statistical basis of diversification.

The result should not be misunderstood. A large portfolio does not eliminate losses or guarantee profit. It improves predictability under appropriate conditions. The insurer still faces estimation error, inflation, changes in exposure, claim severity uncertainty, operational expenses, investment risk and events affecting many policyholders simultaneously.

Independence and correlation

The benefits of pooling are greatest when insured events are independent or only weakly related. A burglary at one unrelated household does not usually cause a burglary at another. Such risks can be diversified effectively. By contrast, an earthquake, flood, pandemic or major cyber incident may generate many claims from the same underlying cause. These risks are correlated, and the law of large numbers provides less protection because the losses move together.

Correlation is a central portfolio-management concern. Two risks may appear to belong to different policyholders but still be exposed to the same peril, supplier, geographic region, infrastructure system or economic factor. A property insurer with thousands of buildings may remain highly concentrated if most buildings are located in one floodplain. A cyber insurer may have a diversified customer list but substantial aggregation if many customers rely on the same cloud service provider.

The law of large numbers

The law of large numbers states, in practical terms, that as the number of comparable and suitably independent observations increases, the average observed outcome tends to approach its expected value. Insurers use this principle to estimate claim frequency, average severity and aggregate loss distributions. Large and credible datasets generally permit more stable estimates than small datasets, although data quality is as important as data volume.

The weak form of the law concerns convergence of averages over repeated observations. The strong form concerns almost sure convergence under specified mathematical conditions. Insurance practice does not depend on reciting these formal distinctions, but it does depend on understanding their assumptions. The risks must be sufficiently comparable, the exposure definition must remain stable, and the underlying probability process must not change radically during the period of observation.

Actuarial pricing therefore begins with expected loss but cannot end there. A simplified premium calculation may be represented as:

Premium = expected claims cost + expenses + risk margin + cost of capital − other adjustments.

Expected claims cost may be estimated as expected frequency multiplied by expected severity. The final premium must also reflect acquisition expenses, administration, reinsurance, taxes, capital requirements, uncertainty and the insurer’s strategic objectives. A premium equal only to the mathematical expected loss would generally be insufficient to operate a solvent insurance business.

Homogeneity and credibility

Pooling works better when risks are sufficiently homogeneous. Homogeneity does not mean that every exposure is identical; it means that the insurer can classify relevant differences and estimate their effects. A portfolio combining a small suburban house, a chemical plant and an offshore platform would not be homogeneous for ordinary property pricing. Classification, underwriting and segmentation make the pool analytically manageable.

Credibility refers to the degree to which an observed body of data deserves reliance in estimating future outcomes. A large, stable portfolio may have high statistical credibility. A new product, a rare peril or a single corporate account may require external data, expert judgment, scenario analysis or a credibility-weighted combination of internal and external information. Professional judgment is necessary, but it must be documented and governed to avoid arbitrary pricing.

Adverse Selection: The Pool Before the Contract

Adverse selection arises when individuals who know more about their risk than the insurer are more likely to seek, retain or purchase extensive insurance. The information asymmetry exists before the contract is formed. If high-risk individuals purchase coverage at rates designed for an average population, the insurer’s portfolio may contain a disproportionate share of high-risk exposures. Premiums then rise, lower-risk customers may exit, and the pool can deteriorate further. This sequence is sometimes called a risk-selection spiral.

For example, people who know that they have serious health conditions may be especially motivated to buy generous health insurance, while healthier individuals may consider the premium unattractive. In motor insurance, drivers with risky habits may have greater demand for comprehensive coverage. In property insurance, owners aware of defects or prior losses may be more eager to obtain coverage than owners of well-maintained properties.

Adverse selection is not the same as fraud. A customer may truthfully provide all information requested and still possess private information that makes the risk more attractive to insure from the customer’s perspective than from the insurer’s. The problem is informational rather than necessarily dishonest.

Insurers respond through several mechanisms:

• risk classification and underwriting;
• proposal forms, inspections and verification;
• deductibles, waiting periods and exclusions;
• policy limits and benefit design;
• experience rating and renewal review;
• group or compulsory insurance arrangements;
• minimum participation requirements; and
• reinsurance and portfolio monitoring.

These responses involve trade-offs. Highly detailed underwriting may improve risk selection but increase transaction costs, delay access to insurance and raise concerns about fairness or discrimination. Regulation may restrict the variables that insurers may use, particularly in personal lines. The managerial challenge is to distinguish legitimate risk differentiation from unfair exclusion.

Moral Hazard: Behaviour After Coverage

Moral hazard occurs when insurance changes behaviour in a way that increases the probability or size of loss because the insured does not bear the full financial consequences. The information problem arises after coverage is obtained. A person with comprehensive motor insurance may drive less cautiously; a firm with property insurance may invest less in loss prevention; or a policyholder may exaggerate a valid claim.

Moral hazard can be classified as ex ante or ex post. Ex ante moral hazard concerns reduced prevention before the loss. Ex post moral hazard concerns behaviour after a loss, such as inflating the claim or failing to mitigate further damage. In health insurance, moral hazard may also involve increased use of services because the marginal cost paid by the insured is reduced, although medical necessity and provider incentives complicate the analysis.

Insurance contracts manage moral hazard through deductibles, copayments, coinsurance, exclusions, warranties, safety conditions, claims investigation, salvage rights, subrogation and policy limits. Risk engineering, loss-control inspections and claims audits also matter. These devices preserve some connection between behaviour and financial consequences.

Moral hazard should not be used as a justification for treating every claimant with suspicion. Excessive controls can delay legitimate claims, damage trust and create unfair burdens for vulnerable policyholders. Effective claims governance combines verification with proportionality, service standards and avenues for appeal.

A Framework for Understanding the Insurance Portfolio

The operation of an insurance portfolio can be understood through five linked questions:

1. What exposures are being pooled? The insurer identifies the insured object, peril, period, geography, limits and applicable conditions.

2. How are the exposures related? Analysts examine independence, concentration, accumulation and common-cause dependencies.

3. What information is available? Underwriting considers data quality, disclosure, credibility and possible asymmetry between insurer and applicant.

4. What incentives does the contract create? Deductibles, limits, exclusions and monitoring influence prevention and claims behaviour.

5. Can the insurer absorb adverse outcomes? Capital, reserves, reinsurance and catastrophe financing support the insurer when actual losses exceed expectations.

This framework connects individual underwriting decisions with enterprise risk management. A policy may be profitable in isolation but undesirable if it increases geographic or peril concentration. Conversely, an apparently high-risk account may improve the portfolio if its losses are weakly correlated with existing exposures and its premium adequately reflects uncertainty.

Reinsurance extends the pooling process. The primary insurer transfers part of its assumed risk to a reinsurer, thereby protecting capital and limiting volatility. Proportional reinsurance shares premiums and losses according to agreed percentages. Non-proportional arrangements, such as excess-of-loss cover, respond when losses exceed a specified retention. Reinsurance does not remove risk; it redistributes it across a wider financial network. The solvency and credit quality of reinsurers therefore matter.

Illustrative Case: Flood Portfolio and the Limits of Pooling

Background

Consider RiverShield Insurance, a fictional regional insurer offering household property insurance in a coastal metropolitan area. For several years, the company expanded its customer base and relied on historical claim data showing relatively stable annual flood losses. Its portfolio contained 80,000 households, and management believed that the large number of policies provided adequate diversification.

Situation and problem

A revised climate and hydrological assessment indicated that a significant proportion of the insured properties were exposed to the same storm-surge system. The portfolio was numerically large but geographically concentrated. In addition, many policyholders had purchased coverage only after receiving local warnings or observing repeated flooding in their neighbourhoods. Some properties had been renovated without adequate elevation or drainage improvements. RiverShield faced both adverse selection and moral hazard, while the correlation of losses meant that the law of large numbers would offer limited protection during a severe event.

Analysis

The company analysed claims by elevation, distance from the coast, drainage basin, building construction, prior loss history and policy inception date. Geographic mapping revealed that 55 per cent of the insured value was located in three connected low-lying zones. Scenario analysis showed that a single severe storm could generate losses several times greater than the historical annual average. The historical data had therefore understated tail risk because it contained few observations of extreme events.

The underwriting review also found that policyholders with previous flood experience were more likely to purchase the broadest coverage, while owners in less exposed areas were increasingly choosing higher deductibles or declining flood extensions. This pattern threatened the balance of the pool. At the claims stage, some customers had not undertaken reasonable temporary protection after receiving warnings, creating questions about mitigation obligations and moral hazard.

Approach and strategy

RiverShield adopted a portfolio strategy rather than relying on policy-by-policy pricing alone. It introduced more refined risk zones, required property-level information for high-value risks, adjusted deductibles according to exposure, and offered premium credits for verified flood-resilience measures. It purchased excess-of-loss reinsurance and established an event-accumulation limit. The company also revised customer communications so that exclusions, waiting periods and mitigation duties were clearly explained.

The insurer used geospatial analytics and catastrophe models as decision-support tools, but management required independent validation and recognised model uncertainty. It did not automatically cancel all high-risk policies. Instead, it combined risk-based pricing with minimum protection standards, government-supported arrangements where available and a review process for vulnerable households.

Outcome

The revised programme reduced concentration, improved the quality of exposure data and increased the insurer’s ability to withstand a severe event. Premiums increased for some high-risk properties, while resilience credits reduced the increase for policyholders who invested in protective measures. The portfolio remained exposed to catastrophe risk, but the exposure was more transparent and supported by capital and reinsurance.

Learning points

The case demonstrates that a large number of policies does not guarantee effective diversification. Common-cause exposure can dominate numerical scale. It also illustrates the difference between adverse selection, which affects who enters or remains in the pool, and moral hazard, which affects behaviour after coverage is obtained. Finally, it shows why underwriting, product design, analytics, customer communication, reinsurance and governance must be managed together.

Limitations, Challenges and Professional Judgement

Model and data risk

Insurers may possess extensive data but still make poor decisions if the data are biased, incomplete, outdated or incorrectly classified. Historical averages can be unreliable when climate conditions, medical practice, technology, legal rules or customer behaviour change. Rare events are especially difficult because experience data contain few observations. Models should therefore be tested through sensitivity analysis, stress testing and alternative assumptions.

Correlated and systemic risks

Pandemics, cyber aggregation, financial crises, climate-related events and supply-chain failures may affect many insureds simultaneously. Such risks challenge traditional assumptions of independence and may exceed available capital or reinsurance capacity. They require scenario analysis, aggregation controls, public-private mechanisms and careful policy wording.

Fairness and access

Risk-based pricing improves actuarial accuracy but can make coverage unaffordable for people in high-risk circumstances. Regulators and insurers must balance actuarial soundness with social objectives, consumer protection and non-discrimination. Transparent explanations are increasingly important when automated systems influence pricing or claims decisions.

Common professional mistakes

Common errors include treating the expected loss as the complete premium, assuming that a large portfolio is automatically diversified, ignoring accumulation across products, relying uncritically on historical data, confusing adverse selection with fraud, and using deductibles without considering their effect on access and claims behaviour. Another mistake is to view underwriting, claims and reinsurance as separate technical functions rather than parts of one risk system.

Governance, Ethics and Regulation

Insurance pooling involves delegated financial power. Insurers decide which risks to accept, how to classify applicants, how to investigate claims and how to allocate capital. These decisions should be governed by clear underwriting authority, documented pricing assumptions, conflicts-of-interest controls, audit trails and escalation procedures.

Consumer protection requires clear disclosure of coverage, exclusions, waiting periods, deductibles and claim requirements. Claims should be handled promptly, consistently and in good faith. Data protection is essential because underwriting and claims processes may use health, financial, location and behavioural information. Automated decision systems should be subject to validation, monitoring for unfair outcomes, human oversight and explainable governance.

Solvency regulation supports the collective promise made to policyholders. Capital and reserve requirements are intended to ensure that insurers can meet obligations under adverse conditions. Regulatory stress testing, own-risk and solvency assessments, reinsurance oversight and catastrophe reporting all reflect the principle that an insurer’s risk cannot be evaluated solely by examining individual policies.

Technology, Analytics and Emerging Developments

Digital technologies are changing the measurement and management of pooled risk. Telematics can record driving behaviour and support usage-based motor insurance. Internet-connected devices may provide information about fire, water or machinery risks. Geospatial systems improve catastrophe accumulation analysis. Machine-learning tools can identify patterns in claims data, detect possible fraud and support triage.

These technologies create benefits but also introduce risks. A predictive model may reproduce historical discrimination, use proxy variables for protected characteristics, or perform poorly when conditions change. Real-time monitoring may raise privacy concerns and alter the balance of information between insurer and insured. Cybersecurity and data governance are therefore part of insurance risk management, not peripheral administrative issues.

Emerging risks are also reshaping the boundaries of insurability. Climate change increases uncertainty about frequency, severity and correlation of natural hazards. Autonomous vehicles may shift liability from drivers toward manufacturers and software providers. Artificial intelligence may create new professional, cyber and liability exposures. Parametric insurance, which pays when an objectively measured trigger such as rainfall or wind speed is reached, can provide rapid relief but may create basis risk because the trigger does not exactly equal the policyholder’s actual loss.

Best Practices for Managing a Risk Pool

A professionally managed pool generally follows these practices:

• define the exposure and coverage trigger precisely;
• classify risks using relevant, lawful and evidence-based variables;
• monitor concentration by geography, peril, customer group, supplier and policy wording;
• distinguish frequency risk from severity and tail risk;
• use deductibles and limits to align incentives without undermining protection;
• combine internal experience with external data and expert judgment;
• test models through stress, scenario and sensitivity analysis;
• maintain adequate reserves, capital and reinsurance;
• monitor adverse selection at acquisition and renewal;
• investigate claims proportionately and protect legitimate customer rights; and
• review assumptions continuously as technology, regulation and environmental conditions change.

Chapter Summary

Insurance operates through the interaction of mutuality, risk transfer and pooling. Policyholders contribute premiums to a common financial arrangement, while the insurer accepts contractual responsibility for specified losses. Diversification and the law of large numbers make aggregate outcomes more predictable when exposures are sufficiently numerous, comparable and independent. They do not eliminate risk, particularly when losses are correlated or catastrophic.

Adverse selection is an information problem before the contract: higher-risk individuals may be more likely to purchase or retain extensive coverage. Moral hazard is an incentive problem after coverage: insurance may reduce prevention or encourage excessive or inflated claims. Underwriting, classification, deductibles, exclusions, monitoring, claims controls, reinsurance and capital management help address these challenges.

The effectiveness of a pool depends on data quality, portfolio composition, contract design, governance and solvency resources. Modern analytics and digital technologies improve risk assessment but create concerns involving model uncertainty, privacy, cybersecurity and fairness. Insurance professionals must therefore combine statistical reasoning with ethical judgment, regulatory awareness and enterprise-level portfolio management.

Key Terms

Adverse selection: The tendency of higher-risk individuals, possessing relevant private information, to seek or retain insurance more actively than lower-risk individuals.

Aggregate loss: The total loss incurred by all insured exposures in a portfolio during a specified period.

Basis risk: The possibility that an insurance trigger or index payment does not match the policyholder’s actual loss.

Correlation: The degree to which losses from different exposures move together.

Credibility: The extent to which available data deserve reliance in estimating future losses.

Deductible: The portion of a covered loss retained by the policyholder before the insurer pays.

Diversification: The reduction of relative portfolio volatility by combining exposures whose outcomes are not perfectly related.

Expected loss: The probability-weighted average financial loss anticipated from an exposure or portfolio.

Independent risks: Risks whose outcomes do not materially influence one another.

Law of large numbers: The statistical principle that average observed outcomes tend toward their expected values as the number of suitable observations increases.

Moral hazard: A change in behaviour after insurance is obtained that increases the likelihood or cost of loss.

Mutuality: The collective bearing of members’ losses through contributions to a common fund.

Pooling: The combination of many exposures so that aggregate loss becomes more predictable than individual loss.

Premium: The payment made for insurance protection, including amounts supporting expected losses, expenses, risk and capital.

Reinsurance: Insurance purchased by an insurer to transfer part of its assumed risk to another insurer.

Risk classification: The grouping of exposures according to characteristics relevant to expected loss and uncertainty.

Risk transfer: The contractual shifting of specified financial consequences of loss from one party to another.

Systematic or common-cause risk: Risk arising from an event or factor capable of affecting many exposures at the same time.

Underwriting: The process of selecting, assessing, pricing and setting terms for insurance risks.

Review Questions

Short Answer Questions

1. What is meant by mutuality in insurance?

2. Why does diversification reduce relative volatility in an insurance portfolio?

3. Distinguish between adverse selection and moral hazard.

4. What is the role of a deductible in controlling insurance incentives?

5. Why does correlation reduce the effectiveness of the law of large numbers?

Long Answer Questions

1. Explain how risk pooling and risk transfer operate together in an insurance contract.

2. Discuss the assumptions and practical limitations of the law of large numbers in insurance.

3. Analyse the causes and consequences of adverse selection and describe the main underwriting responses.

4. Explain ex ante and ex post moral hazard and evaluate the contractual mechanisms used to manage them.

5. Discuss how reinsurance, capital and portfolio monitoring support the stability of an insurer.

Analytical / Application Questions

1. An insurer has 50,000 household policies, but 70 per cent of the insured value lies in one flood-prone district. Assess whether the portfolio is well diversified and recommend controls.

2. A health insurer observes that healthier customers are cancelling comprehensive cover while customers with known illnesses are retaining it. Analyse the market problem and propose a balanced response.

3. A motor insurer introduces telematics-based pricing. Evaluate how the technology could reduce adverse selection and moral hazard, and identify the ethical and governance risks it creates.

22. Multiple Choice Questions with Answers

Question 1

What is the primary economic purpose of pooling many insured risks?

A. To eliminate every possibility of loss

B. To make aggregate losses more predictable

C. To guarantee that each policyholder receives a claim payment

D. To remove the need for underwriting

Correct Answer: B. To make aggregate losses more predictable

Question 2

Which condition most reduces the diversification benefit of an insurance portfolio?

A. A large number of unrelated household risks

B. Similar policy periods

C. High correlation caused by a common flood event

D. Accurate exposure data

Correct Answer: C. High correlation caused by a common flood event

Question 3

Adverse selection primarily occurs because of information asymmetry that exists:

A. After a claim has been paid

B. Before the insurance contract is formed

C. Only between reinsurers

D. Only during investment management

Correct Answer: B. Before the insurance contract is formed

Question 4

Which example best illustrates moral hazard?

A. An applicant knows that a building has hidden structural defects before applying

B. An insurer miscalculates the expected claim frequency

C. A policyholder takes fewer precautions because insurance covers most of the loss

D. Two unrelated risks experience losses independently

Correct Answer: C. A policyholder takes fewer precautions because insurance covers most of the loss

Question 5

Why is a premium based only on expected claims cost generally inadequate?

A. It ignores expenses, risk margin and capital requirements

B. Expected claims cost is always zero

C. It prevents the use of deductibles

D. It guarantees excessive underwriting profit

Correct Answer: A. It ignores expenses, risk margin and capital requirements

Question 6

Which situation is most likely to have high statistical credibility?

A. A new product with no claims history

B. A rare catastrophe observed once

C. A large, stable and comparable portfolio with reliable data

D. A single unusual commercial account

Correct Answer: C. A large, stable and comparable portfolio with reliable data

Question 7

In the RiverShield case, why did a large number of policies fail to provide complete diversification?

A. All policyholders made claims every year

B. The policies had no contractual terms

C. A substantial proportion of exposure was concentrated in the same flood system

D. The insurer did not collect premiums

Correct Answer: C. A substantial proportion of exposure was concentrated in the same flood system

Question 8

Which arrangement transfers part of an insurer’s assumed risk to another insurer?

A. Coinsurance by the policyholder

B. Reinsurance

C. A waiting period

D. Risk classification

Correct Answer: B. Reinsurance

Question 9

What is a potential governance concern when machine learning is used in underwriting?

A. It makes all risks independent

B. It guarantees unbiased pricing

C. It may reproduce historical discrimination or rely on unsuitable proxies

D. It eliminates the need for regulatory oversight

Correct Answer: C. It may reproduce historical discrimination or rely on unsuitable proxies

Question 10

A parametric flood policy pays when rainfall exceeds a specified threshold, even if the customer’s actual damage differs. This illustrates:

A. Moral hazard only

B. Basis risk

C. Complete indemnity

D. Mutual ownership

Correct Answer: B. Basis risk