6. Underwriting and Risk Selection

Introduction and Context

Insurance converts uncertain individual losses into a collectively financed arrangement. This arrangement works only when risks are assessed systematically and premiums bear a reasonable relationship to expected loss, expenses, capital requirements, and uncertainty. Underwriting is the professional process through which an insurer evaluates a proposal, determines whether the risk is acceptable, classifies it, fixes terms and conditions, and monitors the resulting portfolio. Risk selection is the broader managerial activity of choosing, combining, and controlling risks so that the insurer’s portfolio remains financially and operationally sustainable.

Underwriting is therefore more than accepting or rejecting applications. It involves information gathering, verification, classification, pricing, deductibles, limits, exclusions, warranties, reinsurance, authority controls, and post-acceptance review. In life and health insurance, the principal concerns are mortality, morbidity, longevity, medical uncertainty, and the applicant’s financial purpose. In general insurance, attention is directed to physical hazards, legal liability, catastrophe accumulation, business interruption, fraud, and the frequency and severity of claims.

Effective underwriting seeks neither the lowest possible claim cost nor the broadest possible business volume in isolation. Its objective is a balanced portfolio in which risks are adequately priced, fairly treated, legally compliant, and consistent with the insurer’s risk appetite.

Learning Objectives

After studying this chapter, students should be able to:

1. Explain the meaning, purpose, and stages of underwriting and risk selection.

2. Distinguish mortality, morbidity, property, liability, and catastrophe risks.

3. Evaluate medical, financial, occupational, behavioural, geographical, and exposure information.

4. Apply classification, pricing, exclusions, deductibles, limits, and referral decisions.

5. Analyse underwriting authority, portfolio control, reinsurance, and accumulation risk.

6. Identify ethical, regulatory, technological, and governance issues in underwriting.

7. Recommend an underwriting approach in practical insurance scenarios.

The Underwriting Decision: From Proposal to Portfolio

A typical underwriting decision passes through several connected stages. First, the insurer defines the product, target market, appetite, prohibited risks, and available capacity. Second, it collects proposal information and verifies material facts. Third, the risk is analysed using manuals, actuarial models, inspections, medical evidence, financial statements, external databases, and expert judgement. Fourth, the risk is classified into a relevant rating category. Finally, the insurer decides whether to accept, decline, postpone, accept with modified terms, or refer the case to a higher authority.

Acceptance may involve a standard premium or a modified premium. Modification can include an extra mortality or morbidity charge, a higher deductible, a sub-limit, an exclusion, a warranty, a waiting period, a reduced sum insured, a co-insurance requirement, or a special condition. The decision must then be documented, communicated, issued, and reviewed. Underwriting does not end at policy issuance: changes in exposure, claims experience, renewal information, and emerging hazards may require revised terms.

The central principle is materiality. A fact is material when it could influence the insurer’s assessment of the risk, the terms offered, or the decision to accept it. The relevance of a fact depends on the product and jurisdiction. For example, an applicant’s medical history may be central to life insurance but irrelevant to a commercial property policy. Underwriters must distinguish relevant risk information from intrusive or discriminatory information.

Risk Classification and Homogeneity

Insurance relies on grouping broadly similar risks. A classification should produce groups with sufficiently comparable expected loss characteristics. If risks with very different levels of hazard are charged the same premium, low-risk customers may subsidise high-risk customers and may leave the pool. Conversely, excessive segmentation can create complexity, unstable rates, privacy concerns, and unfair treatment.

Classification usually considers both frequency and severity. Frequency is how often losses occur; severity is the size of loss when it occurs. A private motor portfolio, for example, may have frequent small claims, while industrial property may have infrequent but severe fire or catastrophe losses. Underwriting must also consider uncertainty around expected loss and the insurer’s capital requirements.

Adverse Selection and Moral Hazard

Adverse selection arises when people with higher-than-average risk are more likely to seek or retain insurance, especially when the insurer cannot distinguish them adequately. Medical underwriting, waiting periods, deductibles, policy limits, and credible disclosure requirements help control it.

Moral hazard occurs when insurance changes behaviour in a way that increases loss probability or severity. It may be intentional, such as fraud, or unintentional, such as reduced care because a loss is insured. Risk management requirements, inspections, deductibles, claims investigation, warranties, and experience monitoring reduce moral hazard. Underwriters should remember that strict terms can also create customer dissatisfaction and concealment; controls must be proportionate.

Life Insurance Underwriting

Life underwriting estimates the probability and timing of death, while also considering the policy’s financial purpose and contractual design. The decision generally combines medical, financial, occupational, lifestyle, and policy information.

Medical evidence may include the proposal form, physician statements, medical examinations, laboratory tests, prescription history, and prior claims information, subject to law and consent. Relevant factors can include age, build, blood pressure, diabetes control, cardiovascular history, cancer history, respiratory disease, family history, and functional capacity. The underwriter does not simply identify a diagnosis; the task is to assess its severity, treatment, stability, prognosis, and interaction with other factors.

Financial underwriting is particularly important for large sums assured, business insurance, key-person cover, and policies involving income replacement. It examines income, net worth, existing insurance, liabilities, business ownership, and the economic purpose of the proposed benefit. Financial underwriting limits over-insurance, discourages speculative cover, and reduces incentives for misrepresentation or early claims.

Occupational and avocational risks include work at height, mining, aviation, offshore work, explosives, commercial diving, racing, mountaineering, and other hazardous activities. Underwriting may consider frequency of exposure, protective equipment, training, supervision, safety culture, and whether the activity is professional or recreational. Behavioural information such as tobacco use, alcohol or substance misuse, driving history, and adherence to treatment may affect assessment, but its use must comply with applicable law and privacy principles.

A life risk may be accepted at standard rates, accepted with an extra premium, accepted with a temporary or permanent exclusion, postponed pending evidence, or declined. A postponement is appropriate where information is incomplete or a recent medical event makes prognosis uncertain; it should not be used merely as an unexplained substitute for analysis.

Health and Disability Underwriting

Health insurance underwriting focuses on morbidity: the likelihood of illness, medical treatment, hospitalisation, disability, or long-term care. Its complexity arises from changing medical technology, provider behaviour, inflation, benefit design, lifestyle, and the interaction between medical conditions and utilisation patterns.

Important information includes age, prior diagnoses, treatment history, medications, hospital admissions, family history, occupation, geographical access to care, and the scope of benefits requested. Underwriters may assess waiting periods, pre-existing condition provisions, benefit limits, network arrangements, co-payments, deductibles, and exclusions. In disability insurance, the definition of disability is critical. “Own occupation,” “any occupation,” and functional-ability definitions produce different levels of exposure.

Health underwriting must avoid confusing a person’s medical condition with an automatic assumption of poor claims behaviour. Evidence should be relevant, reliable, proportionate, and interpreted in context. Group health schemes often use broader community or experience-based rating, where individual medical underwriting is limited. The insurer then controls risk through plan design, eligibility rules, provider management, utilisation review, and aggregate portfolio analysis.

Property Underwriting

Property underwriting evaluates the possibility that physical assets will be damaged or destroyed and estimates the resulting financial loss. The analysis begins with the property’s construction, occupancy, protection, exposure, and location. Construction includes materials, age, condition, fire resistance, electrical systems, and structural quality. Occupancy concerns how the premises are used and the hazards created by manufacturing, storage, chemicals, cooking, or public access. Protection includes alarms, sprinklers, fire brigade access, security, emergency procedures, and maintenance. Exposure includes neighbouring premises, flood plains, seismic zones, windstorm paths, and surrounding combustible material.

The value at risk must be measured carefully. Underinsurance can lead to inadequate indemnity and average or coinsurance penalties, while overinsurance may create unnecessary premium and moral hazard. Business interruption underwriting requires analysis of gross profit, dependency on suppliers and customers, indemnity periods, critical equipment, and recovery plans.

Inspections are especially useful for complex commercial risks. An inspection report should not be treated as a mechanical checklist; it should connect physical conditions to plausible loss scenarios and risk improvement actions. A property may be accepted only if electrical defects are corrected, combustible stock is separated, sprinklers are installed, or security controls are strengthened.

Liability Underwriting

Liability insurance covers legal obligations arising from injury, property damage, financial loss, professional error, environmental harm, or product defects. Liability risks are difficult because claims may be reported long after the event, legal standards evolve, defence costs can be substantial, and a single judgment may be severe.

Underwriters examine the insured’s business activities, contractual responsibilities, customer profile, geographic operations, safety systems, claims history, quality controls, professional qualifications, financial strength, and use of subcontractors. They also analyse policy triggers, such as occurrence or claims-made coverage, retroactive dates, reporting requirements, defence arrangements, aggregate limits, and deductibles.

Professional indemnity underwriting requires attention to the nature of advice or services, client concentration, contract wording, peer review, documentation, regulatory compliance, and claims-made continuity. Product liability underwriting considers product design, testing, quality assurance, recall procedures, distribution territories, and the potential scale of bodily injury or property damage.

Catastrophe and Accumulation Risk

Catastrophe underwriting addresses events capable of producing many claims from one cause, including earthquake, flood, cyclone, wildfire, terrorism, pandemic-related disruption, or major industrial accidents. A risk that appears acceptable individually may be undesirable when combined with thousands of similar exposures in the same geographic area or supply chain.

Accumulation management requires exposure mapping, hazard zones, scenario analysis, probable maximum loss estimates, catastrophe models, event limits, geographic diversification, and reinsurance. Models estimate possible losses but do not eliminate uncertainty. Their results depend on data quality, assumptions, vulnerability functions, hazard definitions, and the treatment of secondary effects. Underwriters should challenge model outputs with historical evidence, engineering judgement, and stress scenarios.

Portfolio control is achieved through aggregate limits, zonal capacity, facultative or treaty reinsurance, exclusions, deductibles, risk improvement, and deliberate diversification. The relevant question is not only “Is this risk acceptable?” but also “What does accepting this risk do to the portfolio?”

Pricing, Terms, and Authority

Premium is commonly developed from expected loss, expenses, cost of capital, reinsurance cost, profit provision, and uncertainty margins. Rating may use manual rates, experience rating, exposure rating, or a combination. Manual rating applies a tariff or underwriting schedule to a class. Experience rating gives weight to the insured’s own claims experience where credible. Exposure rating estimates expected loss from measurable exposure units such as payroll, revenue, vehicle count, property value, or lives insured.

Underwriters should not use price as the only response to uncertainty. If a hazard cannot be priced reliably, an exclusion, limit, deductible, inspection requirement, or decline may be more appropriate. Conversely, excessive exclusions can undermine the product’s purpose and create disputes.

Delegated authority systems specify the classes, limits, conditions, and premium ranges within which an underwriter may decide. Referral rules are necessary for large sums insured, unusual occupations, severe medical impairment, hazardous materials, poor claims experience, sanctions concerns, or catastrophe accumulation. Governance requires audit trails, peer review, exception reporting, and separation of sales incentives from technical decision-making.

An Illustrative Case: Meridian Food Processors

Background

Meridian Food Processors, a medium-sized manufacturer, seeks property, business interruption, product liability, and employers’ liability insurance. Its main plant is located near a river and stores large quantities of combustible packaging. The company has recently expanded production and installed automated refrigeration equipment.

Situation and Problem

The proposal reports no major losses during the previous five years. However, the inspection identifies inadequate separation between raw materials and finished goods, an intermittently functioning sprinkler system, dependence on one refrigeration supplier, and flood exposure. Revenue has increased sharply, but the requested business interruption sum insured is based on last year’s figures. A product recall occurred in another country involving a similar product category.

Analysis

The absence of recent claims does not establish a low-risk profile. Physical protection is deficient, flood could affect both property and machinery, and supplier concentration increases interruption severity. The business interruption value is potentially inadequate. Product liability requires analysis of quality controls, traceability, batch testing, and distribution territories. The combined exposure also creates an accumulation issue if the insurer already covers other businesses in the same flood zone.

Approach

The underwriter refers the account to senior authority, requests updated financial and exposure data, and obtains engineering recommendations. Acceptance is made conditional on sprinkler repair, stock segregation, electrical maintenance, and a documented flood response plan. A flood deductible and sub-limit are applied, subject to portfolio capacity. The business interruption sum insured and indemnity period are recalculated. Product liability is offered with an aggregate limit, a clear territorial scope, recall-related wording, and evidence of quality assurance. Facultative reinsurance is arranged for the portion exceeding the insurer’s preferred capacity.

Outcome

The account is accepted on modified terms rather than declined. The insured improves its protection systems and revises its continuity plan. The insurer obtains a more credible exposure base, controls accumulation, and aligns terms with the actual hazard. The case demonstrates that underwriting is a negotiated risk-management process, not merely a pricing exercise.

Learning Points

1. Claims history must be interpreted with physical and operational evidence.

2. Exposure values and indemnity periods require regular updating.

3. Conditions can reduce risk, but they must be specific, measurable, and monitored.

4. Individual acceptance decisions must be tested against portfolio accumulation.

5. Reinsurance expands capacity but does not replace sound primary underwriting.

Professional and Ethical Dimensions

Underwriters exercise significant influence over access to financial protection. Fair treatment requires consistent criteria, understandable communication, appropriate disclosure standards, and a documented basis for adverse decisions. Anti-discrimination rules may restrict the use of certain personal characteristics or require evidence that a rating factor is actuarially relevant and legally permitted.

Privacy is central because underwriting involves sensitive medical, financial, behavioural, and location data. Information should be collected for a legitimate purpose, used proportionately, protected from unauthorised access, and retained only as long as necessary. Consent, data accuracy, correction rights, and secure sharing with reinsurers or service providers require governance.

Conflicts may arise when sales targets encourage rapid acceptance, when delegated partners receive incentives linked to volume, or when claims data are used without adequate explanation. Strong governance includes underwriting manuals, model validation, complaints processes, conflict disclosures, quality assurance, conduct monitoring, and regulatory reporting.

Technology, Analytics, and Emerging Practice

Digital underwriting increasingly uses electronic proposals, automated rule engines, geospatial information, telematics, wearable data, medical databases, remote inspections, and natural-language processing. Automation can improve speed and consistency, while predictive analytics may identify hidden relationships in claims and exposure data.

These tools create risks of biased data, opaque decisions, cyberattack, excessive surveillance, and model drift. A model trained on historically unequal outcomes may reproduce those outcomes. Human oversight is therefore essential for complex, vulnerable, or adverse cases. Underwriters should be able to explain the principal factors influencing a decision and challenge an implausible automated result.

Climate change, supply-chain interdependence, cyber exposures, ageing populations, antimicrobial resistance, and changing work patterns are reshaping risk selection. Historical averages may be less reliable where hazard frequency or severity is changing. Scenario analysis, forward-looking exposure monitoring, climate-sensitive pricing, and resilient risk improvement are increasingly important.

Common Errors and Best Practices

Frequent underwriting errors include relying on incomplete proposal forms, accepting sales descriptions without verification, ignoring accumulation, using outdated values, treating models as unquestionable, failing to document deviations, and applying exclusions so broadly that coverage becomes illusory. Another error is confusing a clean claims record with a safe risk; a short observation period may simply have produced no loss yet.

Best practice is to:

• define appetite and referral thresholds clearly;
• obtain decision-relevant information rather than unnecessary personal data;
• combine statistical evidence with professional judgement;
• distinguish frequency, severity, uncertainty, and accumulation;
• use terms that are proportionate, clear, and enforceable;
• review exposure and claims information at renewal;
• test pricing and models under adverse scenarios;
• maintain audit trails and independent quality review; and
• integrate underwriting with claims, actuarial, risk engineering, compliance, and reinsurance functions.

Chapter Summary

Underwriting and risk selection determine how an insurer converts uncertain exposures into manageable portfolios. The process includes information collection, verification, classification, pricing, terms, authority, issuance, and continuing review. Life underwriting emphasises mortality, financial purpose, medical evidence, occupation, and lifestyle. Health underwriting focuses on morbidity, benefit design, utilisation, and disability definitions. Property underwriting examines construction, occupancy, protection, exposure, values, and business interruption. Liability underwriting considers legal responsibility, claims development, contracts, safety, and limits. Catastrophe underwriting adds accumulation, modelling, geographic concentration, scenarios, and reinsurance.

Sound decisions balance expected loss, uncertainty, capital, fairness, customer value, and portfolio strategy. Technology can improve consistency and insight, but privacy, bias, explainability, and human oversight remain essential. The professional underwriter combines evidence and judgement, documents decisions, manages exceptions, and treats each risk as part of a wider portfolio.

Key Terms

Adverse selection: The tendency for higher-risk persons to seek or retain insurance disproportionately.

Accumulation risk: Concentration of multiple insured exposures capable of being affected by one event.

Business interruption: Loss of income or additional expense resulting from interruption of insured operations.

Catastrophe model: A quantitative model estimating losses from defined hazardous events.

Deductible: The portion of an insured loss borne by the policyholder.

Exposure: The person, property, activity, value, or circumstance subject to possible loss.

Experience rating: Pricing influenced substantially by the insured’s credible claims history.

Financial underwriting: Assessment of the economic purpose and adequacy of proposed insurance.

Hazard: A condition that increases the probability or severity of loss.

Indemnity period: The period for which business interruption loss may be covered.

Material fact: Information that could influence acceptance, terms, or pricing.

Morbidity: The incidence or likelihood of illness, disability, or medical treatment.

Mortality: The incidence or likelihood of death within a specified population.

Moral hazard: Increased loss potential arising from behaviour altered by insurance.

Portfolio risk: The aggregate risk arising from the insurer’s collection of policies.

Risk appetite: The types and amount of risk an insurer is prepared to accept.

Underwriting: Evaluation and selection of risks and determination of insurance terms.

Waiting period: A specified period before selected benefits become payable.

Review Questions

#### Short Answer Questions

1. What is the relationship between underwriting and risk selection?

2. Define adverse selection and state one control used against it.

3. Why is financial underwriting important in life insurance?

4. Identify the four traditional property underwriting considerations.

5. What is accumulation risk?

#### Long Answer Questions

1. Explain the stages of the underwriting process from proposal to portfolio review.

2. Compare mortality underwriting with morbidity underwriting.

3. Discuss the principal factors used in commercial property underwriting.

4. Explain how liability underwriting differs from property underwriting.

5. Analyse the role of reinsurance, limits, deductibles, and exclusions in catastrophe risk control.

#### Analytical / Application Questions

1. A life applicant requests a large policy but reports modest income and substantial existing cover. What information should the underwriter seek, and what decisions might follow?

2. A warehouse has excellent recent claims experience but is located in a flood zone with inadequate sprinklers. Should it be accepted at standard terms? Justify your answer.

3. An automated underwriting model declines applicants from a particular geographic area at unusually high rates. How should the insurer investigate and govern this outcome?

22. Multiple Choice Questions with Answers

Question 1

Which statement best describes the central purpose of underwriting?

A. To maximise the number of policies issued regardless of risk

B. To evaluate, classify, price, and control risks within the insurer’s appetite

C. To settle claims before policy issuance

D. To eliminate all uncertainty from insurance portfolios

Correct Answer: B. To evaluate, classify, price, and control risks within the insurer’s appetite

Question 2

A life insurer requests income details and existing insurance for a very large sum assured primarily to perform:

A. Geographic accumulation analysis

B. Financial underwriting

C. Property protection analysis

D. Provider utilisation review

Correct Answer: B. Financial underwriting

Question 3

Which situation most directly illustrates adverse selection?

A. An insured installs sprinklers after an inspection

B. A person with concealed serious illness is especially eager to purchase cover

C. A reinsurer diversifies its geographic portfolio

D. A policyholder pays a deductible after a loss

Correct Answer: B. A person with concealed serious illness is especially eager to purchase cover

Question 4

In commercial property underwriting, construction, occupancy, protection, and exposure are used mainly to assess:

A. The legal enforceability of a claims-made policy

B. Physical property hazard and potential loss

C. The applicant’s mortality rate

D. The credibility of a life table

Correct Answer: B. Physical property hazard and potential loss

Question 5

Why might an underwriter impose a flood sub-limit rather than simply charge a higher premium?

A. The exposure may be difficult to price reliably or may exceed desired capacity

B. Premium is never relevant to catastrophe risk

C. Sub-limits automatically eliminate flood risk

D. Flood losses are always small and frequent

Correct Answer: A. The exposure may be difficult to price reliably or may exceed desired capacity

Question 6

Which factor is particularly important in liability underwriting?

A. Only the age of the building

B. The insured’s legal responsibilities, activities, contracts, and claims development

C. The applicant’s blood pressure

D. The number of lives in a mortality table

Correct Answer: B. The insured’s legal responsibilities, activities, contracts, and claims development

Question 7

A clean five-year claims record should not by itself establish that a risk is safe because:

A. Claims records are never useful

B. Physical hazards and future severity may remain significant despite no recent loss

C. Underwriters must always decline risks without claims

D. Claims history applies only to life insurance

Correct Answer: B. Physical hazards and future severity may remain significant despite no recent loss

Question 8

Which control most directly addresses catastrophe accumulation?

A. Ignoring geographic location

B. Using exposure maps, aggregate limits, scenarios, and reinsurance

C. Removing all deductibles

D. Relying only on the applicant’s description

Correct Answer: B. Using exposure maps, aggregate limits, scenarios, and reinsurance

Question 9

What is a principal governance risk of automated underwriting?

A. It makes data protection unnecessary

B. It may reproduce biased historical data or produce unexplained decisions

C. It guarantees that all risks are fairly priced

D. It prevents portfolio diversification

Correct Answer: B. It may reproduce biased historical data or produce unexplained decisions

Question 10

In the Meridian Food Processors case, acceptance on modified terms was appropriate because:

A. The absence of recent losses proved standard terms were adequate

B. Risk improvements, revised values, limits, and reinsurance could address identified hazards

C. The insurer ignored flood exposure

D. Product liability was unrelated to quality controls

Correct Answer: B. Risk improvements, revised values, limits, and reinsurance could address identified hazards