How do I choose a risk management response when a loss is possible?

Identify the asset, activity or legal duty exposed to loss; estimate frequency and severity; then compare avoidance, prevention, reduction, retention and transfer. A frequent, modest loss may justify controls and a deductible, while a rare loss capable of threatening the organization may call for insurance or another financing arrangement. Check the remaining risk after each measure, including exclusions, limits and the organization's ability to fund retained losses.

Why can an insurer report an underwriting loss yet remain profitable?

Underwriting compares earned premium with incurred losses and underwriting expenses. For example, $100 million of earned premium, $72 million of incurred losses and $32 million of expenses produce a $4 million underwriting loss and a 104% combined ratio. If the insurer earns $7 million of net investment income and has no other relevant items in this simplified example, the result before tax is a $3 million profit. The investment result does not erase the need to assess whether pricing and claims costs are sustainable.

What should I establish before deciding whether a policy responds to a claim?

Identify the insured, the claimant's interest, the event date and location, and the alleged injury or damaged property. Then follow the applicable policy's insuring agreement, definitions, exclusions, exceptions, conditions, limits and deductibles in that order, using the law of the stated jurisdiction where it affects interpretation. A covered event may still produce a smaller payment because of a limit, deductible or valuation term; a payment calculation cannot establish coverage by itself.

CPCU Study Guide: Risk, Insurance and Decisions

Start with concept 1 Choose a topic

The CPCU designation brings several kinds of insurance judgment together. You need to identify loss exposures, decide how to manage them, understand how insurer departments act on those decisions, and explain the financial result. You also need to read coverage and legal facts carefully: a loss may be real but fall outside a particular policy, while a sound underwriting decision may still fail if its data are weak. This guide builds those skills across the five core CPCU courses, the commercial and personal concentration alternatives, selected elective alternatives, and Ethics 311. Each planned lesson starts with a decision rule and ends with a resolved situation, so you can see how a change in facts changes the answer. United States legal examples identify the relevant jurisdiction and assumed policy wording where those details control the result.

CPCU 500: Becoming a Leader in Risk Management and Insurance

1. Classify the risk before choosing a remedy

Before choosing insurance or another remedy, identify what can be lost and how. A pure loss exposure has two relevant outcomes: loss or no loss. Fire can damage a warehouse, and an employee can be injured; neither event creates a gain from the event itself. A speculative decision, such as expanding inventory to meet uncertain demand, can produce a gain or a loss. That decision calls for business analysis as well as protection against its pure loss exposures. Describe each exposure by the person, property, income stream or liability affected, then identify the cause and possible financial consequence. This prevents a property policy from being treated as a solution to lost sales caused solely by weak demand. The CPCU 500 course develops this risk analysis; the CPCU 500 study guide provides related practice.

Worked example: A warehouse operator lists three concerns. Fire could destroy $500,000 of stock, and a worker could suffer an injury. Both are pure loss exposures: the event can cause loss, but offers no gain. Demand for a new product could raise or lower profit, so it is speculative. The operator should address fire protection and employee safety while evaluating the product decision through forecasts and risk appetite.

Watch the distinction: Calling every uncertain outcome an insurable loss blurs the decision. Insurance may finance specified fire or injury losses; it does not normally compensate a business simply because demand disappoints.

Reference: [4]

2. Measure frequency and severity together

Frequency describes how often losses occur; severity describes their size. For planning, expected annual loss is estimated event frequency multiplied by average loss severity, or event probability multiplied by loss amount for a single binary event. It is useful for comparing routine costs, but it is only an average across many possible years. A rare event can have a lower expected loss yet threaten the organization's ability to continue operating. Assess the largest plausible loss, cash available after it, and how quickly operations could recover. Also test whether the estimate rests on credible exposure data: twenty thefts at one location may not predict another location's experience. A sound recommendation can retain predictable small losses while using controls and risk financing for a severe event.

Worked example: Suppose a warehouse expects 20 thefts a year at $2,000 each: 20 × $2,000 = $40,000 expected annual loss. A fire has a hypothetical 1% annual probability and would cost $2 million: 0.01 × $2 million = $20,000. Theft has the higher expected loss, but one fire could require $2 million at once. Budget for theft and plan separately for fire protection and financing.

Watch the distinction: Choosing treatment solely by the larger expected loss ignores the size of a single event. A $20,000 annual average does not mean a fire would cost only $20,000.

Reference: [4]

3. Build a layered risk treatment

Risk treatment works best as a sequence. Avoidance stops an activity that creates an unacceptable exposure. Prevention lowers the chance of a loss, while reduction limits damage after an event begins; sprinklers primarily reduce fire severity, though their presence may also change other outcomes. Retention means the organization funds some loss itself, deliberately through a deductible or unintentionally through an uncovered gap. Insurance transfers specified financial consequences above the retained amount, subject to the policy's limits and terms. These measures can coexist. First describe the untreated exposure, then estimate how controls change frequency or severity, identify what the organization can pay, and insure the residual loss that could threaten its objectives. Revisit the arrangement when property values, operations or available cash change.

Worked example: A firm keeps its warehouse, installs sprinklers, and buys hypothetical fire coverage with a $25,000 deductible and enough limit for the loss. On a covered $200,000 fire, the firm pays $25,000 and the insurer pays $200,000 − $25,000 = $175,000. Sprinklers may have reduced the damage, but they do not eliminate the firm's retained amount or every possible fire loss.

Watch the distinction: Buying insurance does not remove the physical risk. Controls still protect people and operations, while deductibles, exclusions and limits can leave the organization with substantial financial exposure.

Reference: [4]

4. Defend a strategic recommendation

A strategic recommendation should state the decision, the evidence supporting it, and what could make it wrong. Compare the total cost of each option with its expected change in losses, but distinguish an estimated saving from a guaranteed one. Check whether the data represent the same operations and whether implementation could disrupt work. Then consider risk appetite: an option with a small expected saving may still be attractive if it reduces severe injuries, regulatory exposure or downtime. Different stakeholders will see different effects. Finance needs the cash cost and uncertainty; operations needs a workable rollout; employees need to understand the safety change. Present a measurable review point, such as claims frequency and severity after implementation, so the organization can update the decision when actual results arrive.

Worked example: A proposed safety program costs $60,000 a year. Comparable exposure data suggest annual claims will fall from $140,000 to $65,000, a reduction of $75,000. Expected net saving is $75,000 − $60,000 = $15,000 a year. Recommend a monitored rollout if the comparison is credible, and explain that weaker loss reduction could erase the saving even if the program still improves safety.

Watch the distinction: A positive expected net saving is not proof that every year will be profitable. Claims vary, and an unreliable baseline can make the forecast look stronger than it is.

Reference: [4]

CPCU 520: Meeting Challenges Across Insurance Operations

5. Trace the insurance value chain

An insurer's value chain turns information about an exposure into a priced promise, then fulfills that promise when a covered loss occurs. Distribution identifies customers and gathers applications. Underwriting decides whether the risk fits the insurer's strategy and what terms to offer. Pricing estimates the premium needed for expected claims, expenses and other financial needs. Risk control can reveal hazards and improve them. Claims tests coverage, measures damage and settles supported losses; finance tracks the resulting obligations and capital. Information must travel in both directions. A claim pattern can reveal an underwriting question that was missed, while a risk control finding can change a quote before binding. If one function treats incomplete data as reliable, later functions may inherit the error.

Worked example: A producer seeks a warehouse quote but leaves construction type blank. Underwriting asks for the missing fact rather than assuming a fire resistance level. Risk control confirms working sprinklers but finds combustible storage near a heat source. Pricing uses the verified construction and controls to revise the proposed terms. If the account is accepted, claims later records any fire cause so underwriting can assess similar risks.

Watch the distinction: A sale is not the end of the value chain. Poor application data can distort pricing, and claim findings can expose patterns that require changes in underwriting or risk control.

Reference: [5]

6. Connect claims work to the insurer's promise

Claims work fulfills the insurer's promise through several distinct decisions. Investigation establishes what happened, the amount of damage and the evidence supporting each fact. Coverage analysis applies the policy grant, exclusions, conditions, limits and deductible to those facts. Reserving estimates the insurer's unpaid obligation as information changes; it is an estimate, not a settlement offer. Settlement communicates and pays the supported covered amount, or explains a partial or full denial. Separate gross reported damage from covered damage and from the insurer's net payment. A claim file should show why those figures differ, including estimates, photographs and applicable terms. Patterns in causes of loss can inform underwriting, pricing and risk control, provided the data distinguish true exposure changes from changes in claim handling.

Worked example: A storm claim is reported at $90,000. Inspection supports $60,000 in storm damage; the other $30,000 reflects earlier deterioration under the assumed policy. With no deductible or prior payment in this hypothetical, the adjuster sets the unpaid reserve at $60,000, explains the excluded portion, and settles $60,000 when documentation is complete. The storm cause is shared with underwriting for portfolio review.

Watch the distinction: The first amount reported is not automatically the reserve or payment. Investigation can change both damage estimates and coverage, and a reserve remains provisional until the claim develops.

Reference: [5]

7. Coordinate underwriting and distribution

Distribution creates access to customers, but speed must be matched to reliable risk selection. An insurer can define its target market, required application fields, automatic decision rules and referral triggers. Straightforward personal risks with complete data may fit consistent automated rules; unusual commercial operations often need judgment about processes, controls and potential severity. Producers help by gathering accurate facts and explaining customer needs, while underwriters decide whether the proposed terms match the insurer's appetite and capacity. A referral is useful when a material fact is missing, a risk falls outside a rule, or several individually acceptable features create an unusual combination. Measure the system by quote quality, conversion, claim performance and customer fit, not speed alone.

Worked example: A distributor asks for instant quotes on every applicant. The insurer allows automatic quotes when required location, construction, use and loss history fields are complete and within stated rules. A standard home may qualify. A manufacturer with spray finishing is referred because that process changes fire and liability exposures. The producer collects process details; the underwriter then decides on terms or declines.

Watch the distinction: Treating every referral as a failure of automation encourages guesses about missing facts. A fast quote based on incomplete exposure data may be inaccurate for both customer and insurer.

Reference: [5]

8. Link pricing to capacity

Actuarial ratemaking starts with projected losses and expenses for a defined exposure base and period. Divide the total by policy count only when policies are comparable enough for an average to be meaningful. Real rate indications also consider claim development, trends, exposure differences, expenses, risk margin and the insurer's investment or capital assumptions. Underwriting then applies approved classification and individual risk information to an account. Capacity is a separate question: can the insurer absorb a large loss or concentration of losses while meeting its obligations? Reinsurance can reduce the insurer's net retained loss under a specified agreement, allowing it to manage severe exposures or portfolio volatility. Reinsurance has a cost and does not make inadequate underlying pricing adequate by itself.

Worked example: Assume projected annual losses are $8 million and expenses are $2 million for 10,000 comparable one year policies. The simplified break even total is $10 million, so indicated average premium is $10 million ÷ 10,000 = $1,000 per policy before other adjustments. A possible multimillion dollar loss still needs a capacity decision; reinsurance can limit retained severity but its premium must enter a fuller pricing analysis.

Watch the distinction: The $1,000 average is not a final price for every policy. Different exposures need different rates, and buying reinsurance does not erase its cost or repair a weak loss estimate.

Reference: [5]

CPCU 530: Applying Legal Concepts to Insurance

9. Read the insurance contract in sequence

For an assumed United States property policy, start with the declarations: who is insured, what property and locations are listed, the period, limits and deductible. Then read the insuring agreement to identify the covered property, cause and required type of loss. Definitions can change the ordinary meaning of a term. Next test each applicable exclusion, including any exception that restores limited coverage. Finally apply conditions such as notice, proof of loss or valuation duties. Keep separate causes and damaged items separate where the facts permit. This order prevents an exclusion from being used before establishing whether the loss entered the coverage grant. State law can affect interpretation, but the concrete coverage analysis begins with the actual contract language. The CPCU 530 legal concepts guide offers related applications.

Worked example: Assume a United States commercial policy lists a store, covers its stock against direct fire damage, excludes flood, and has sufficient limits with no deductible. A fire causes $40,000 of stock damage; a separate flood damages another room's $15,000 of stock. The fire loss enters the grant and is payable at $40,000. The flood loss enters the initial property description but is removed by the exclusion.

Watch the distinction: Seeing insured property in the declarations does not mean every cause of damage is covered. The cause of loss and exclusions must be tested for each distinct damaged item.

Reference: [6]

10. Identify the property interest

Insurable interest is a financial stake in preserving property. Ownership is the clearest example, but possession and a contractual duty to repair can also create exposure. Identify each asset, who owns it, who must bear a loss under the lease or other agreement, and how much that party could actually lose. A tenant may own equipment outright while also being responsible for improvements installed in leased space. The landlord may have a separate interest in the building. These interests can overlap, so insurance should be coordinated to avoid gaps and disputes over payment, but a claimant should not profit beyond its covered financial loss. Policy definitions, valuation terms and applicable United States state law determine how a particular interest is insured and measured.

Worked example: A tenant's stated lease makes it responsible for replacing $15,000 of installed improvements, and it owns $40,000 of equipment. A covered fire destroys both. Its potential financial exposure is $40,000 + $15,000 = $55,000, assuming the values equal its loss. Equipment ownership and the lease repair duty provide different bases for interest. Payment still depends on the tenant's covered property categories, valuation and limits.

Watch the distinction: A tenant does not lack insurable interest merely because the landlord owns the building. The tenant can lose its own property and may owe for improvements under the lease.

Reference: [6]

11. Work through a tort claim

A negligence analysis asks four linked questions under the applicable United States state law. Did the defendant owe this claimant a duty of reasonable care? Did conduct fall short of that duty? Did the breach actually and legally cause the injury? What damages are supported? Evidence should be assigned to the element it proves. A warning about a spill can support notice and breach, but it does not by itself prove that the spill caused a particular fall or that every claimed expense resulted from it. Other liability theories may use different elements, such as liability based on a defective product or a contractual promise, so do not force every claim into negligence. After identifying the theory, evaluate defenses and the available insurance coverage separately from the underlying legal liability.

Worked example: Assume state law requires a shop to use reasonable care toward visitors. Staff record a spill, leave it unattended, and a visitor falls there. The record supports notice and breach; video and medical evidence must connect the fall to the spill and the injury. The visitor has $12,000 in documented medical bills. That proves a potential damages component, but total liability still depends on causation, other damages and any defenses.

Watch the distinction: An injury on business premises does not establish negligence automatically. The claimant must connect a duty and breach to the injury and prove the resulting damages.

Reference: [6]

12. Assess authority and dispute routes

Actual authority comes from what the insurer expressly or implicitly permits an agent to do. Apparent authority concerns what a third party reasonably believes the agent can do because of the insurer's own representations, not simply because the agent claims power. When a producer makes a disputed promise, identify the written authority, the insurer's outward conduct, what the customer saw, and whether reliance was reasonable under the assumed state law. The distinction matters because an unauthorized statement may still create a dispute about the insurer's responsibility. Preserve communications and the issued contract before choosing a response. A factual misunderstanding may be resolved by explanation or negotiation; mediation can help parties reach agreement, while litigation may be needed when legal rights or material facts remain contested.

Worked example: A producer promises flood coverage despite written instructions barring that promise. The insurer's website lists the producer as authorized to sell its property policies, but says nothing about authority to change terms. Actual authority is absent. Apparent authority requires further facts about the insurer's representations and the customer's reasonable reliance under assumed state law. Gather the quote and messages, then explore a documented negotiated resolution before litigation.

Watch the distinction: A producer cannot create apparent authority merely by announcing it. The analysis turns on the insurer's conduct toward the customer and reasonable reliance under the relevant law.

Reference: [6]

CPCU 540: Contributing to Insurer Financial Performance

13. Read insurer accounts with their purpose in mind

Insurer financial statements answer different questions depending on the accounting basis. United States statutory accounting emphasizes whether an insurer can meet policyholder obligations, so assets that cannot readily support those obligations may be nonadmitted. Statutory surplus equals admitted assets minus liabilities and serves as a capital cushion. Generally accepted accounting principles, or GAAP, aim to present financial performance and position to a broader set of users under different recognition and measurement rules. A surplus change can come from underwriting, investment results, capital transactions or accounting adjustments; it should not be described as profit without reconciliation. Check asset quality and liability estimates as well as the total. A large reported asset balance may offer little claims paying support if material assets are not admitted. The CPCU 540 finance guide develops these comparisons.

Worked example: A simplified United States statutory statement reports $120 million in admitted assets and $92 million in liabilities. Statutory surplus is $120 million − $92 million = $28 million. Suppose the insurer also owns a $3 million asset classified as nonadmitted for this example. Counting it would overstate statutory surplus: the relevant figure remains $28 million, though the asset may appear differently under another accounting basis.

Watch the distinction: Assets minus liabilities gives statutory surplus only when the asset figure uses admitted assets. Including a nonadmitted item would make the solvency cushion appear larger than reported.

Reference: [7]

14. Reconcile premium and claim measures

Written premium records the amount associated with policies issued during a period; earned premium allocates that amount to coverage already provided. The unearned portion represents coverage still owed. For a simple policy with uniform risk over time, earning can be proportional to elapsed time, though actual accounting can require adjustments. Paid losses are cash already disbursed. Incurred losses add the change in estimated unpaid claim liabilities to payments during the period: paid losses plus ending loss reserves minus beginning loss reserves, with other adjustments where applicable. This means incurred loss can rise without a payment when a reserve increases, or fall when an estimate is released. Use earned premium and incurred loss for an underwriting period comparison, while checking that both cover the same business and dates.

Worked example: Assume a $1,200 annual policy starts July 1 and risk earns evenly. At December 31, six of twelve months have elapsed, so $1,200 × 6/12 = $600 is earned and $600 remains unearned. Separately, a claim has $70,000 paid during the period and a $30,000 ending reserve, with no beginning reserve. Incurred loss is $70,000 + $30,000 − $0 = $100,000.

Watch the distinction: Paid loss is not the same as incurred loss. An unpaid claim estimate affects the period's incurred result even though no cash has yet left the insurer.

Reference: [7]

15. Diagnose underwriting ratios

The loss ratio compares incurred losses and loss adjustment expenses with earned premium when those items are included in the chosen definition. The underwriting expense ratio compares underwriting expenses with an appropriate premium denominator; definitions vary, so verify that the numerator and denominator are consistent before combining ratios. For a simplified earned premium illustration, divide each cost category by the same earned premium and add the results. A combined ratio below 100% indicates an underwriting gain under those assumptions; above 100% indicates an underwriting loss. It excludes investment income and may omit other items, so it is not a complete profitability measure. Diagnose changes by separating claim frequency, severity, expense growth, pricing and business mix. The same combined ratio can arise from very different underlying problems.

Worked example: Assume $10 million earned premium, $6.5 million incurred losses including relevant claim costs, and $3 million underwriting expenses, all for the same year and book. Loss ratio = $6.5 million ÷ $10 million = 65%. Expense ratio = $3 million ÷ $10 million = 30%. Combined ratio = 95%, leaving a simplified underwriting margin of 5%, or $500,000, before other results.

Watch the distinction: A 95% combined ratio is not a 5% return on invested capital. It describes a simplified underwriting result relative to premium, without investment income or capital employed.

Reference: [7]

16. Balance investments and capital

An insurer invests premiums while preparing to pay claims, so asset selection must match the timing and uncertainty of liabilities. Higher yield can come with price volatility, credit risk or difficulty selling an asset promptly. Liquidity matters when claims must be paid before investments mature. Start with expected near term payments and a margin for uncertainty, then hold assets that can supply cash when needed. The remaining portfolio can be evaluated for yield and risk within the insurer's capital capacity. Rapid premium growth can require more capital because new policies create additional possible losses before their outcomes are known. Catastrophe concentrations can produce many claims together, increasing the need for accessible funds and loss absorbing surplus. Investment return cannot safely replace an assessment of claim obligations.

Worked example: An insurer expects $20 million of claim payments over the next quarter. Assume it has $25 million available to invest and no other usable cash. Placing all $25 million in an asset that cannot be sold for a year would leave a $20 million payment gap. It should hold at least the expected $20 million in suitably liquid assets, plus a prudent uncertainty allowance, before considering the illiquid asset.

Watch the distinction: The highest stated yield is not automatically the best insurer investment. A profitable asset that cannot fund due claims may force costly borrowing or sales elsewhere.

Reference: [7]

CPCU 550: Maximizing Value with Data and Technology

17. Test whether data answer the question

Start an analysis by defining the decision and the unit of observation: claim, policy, building or insured. Then check where each field came from, whether its meaning stayed constant, and which records are missing. A recorded claim date might mean loss date, report date or entry date; mixing them can manufacture a trend. Missingness can also be informative. If older claims lack dates more often, discarding incomplete records may make recent claims appear unusually numerous. Compare the retained and excluded groups by age, product and claim size before interpreting results. Representativeness matters when a model or trend will be used in a different population, such as applying commercial property experience to homeowners. The CPCU 550 course covers data use across insurer decisions.

Worked example: A file contains 1,000 claims; 200 lack loss dates, leaving 800 usable for a date trend. Completeness is 800 ÷ 1,000 = 80%. Suppose 150 of the missing dates belong to older claims. A chart built only from the 800 dated records will undercount older years disproportionately, so an apparent rise toward recent years may be a recording artifact. Recover dates or qualify the trend.

Watch the distinction: An 80% complete file is not necessarily an 80% accurate trend. Bias depends on which records are missing and whether their omission changes the comparison.

Reference: [8]

18. Interpret a risk model

A risk model ranks or classifies exposures using patterns in observed data. Its output predicts an outcome under specified conditions; a predictive variable need not cause the loss. Judge performance on buildings like those where the model will operate and over the same forecast period. Precision asks what fraction of flagged buildings actually have a loss. Recall asks what fraction of all buildings with losses were flagged. These measures answer different operational questions. Low precision can waste inspection capacity, while low recall can miss preventable losses. Compare the cost of a false alarm with the cost of a missed loss, and check whether flags are concentrated in a group because of data quality or a proxy variable. Use the model to support a proportionate decision, then monitor results.

Worked example: A model flags 100 buildings; 20 later have losses. Ten more losses occur among 900 unflagged buildings. Precision is 20 ÷ 100 = 20%, and recall is 20 ÷ 30 = 66.7% approximately. The flagged loss rate exceeds the unflagged rate of 10 ÷ 900, about 1.1%, yet 80 flags are false alarms. Use inspection costs and potential loss severity before acting on every flag.

Watch the distinction: A strong difference in loss rates does not prove that the model's inputs cause losses. It also does not make every flagged building a good candidate for costly intervention.

Reference: [8]

19. Place technology where it helps

Automation fits tasks with reliable inputs, repeatable rules and clear exceptions. A claims system can check required fields, route documents, calculate a straightforward amount under stated terms and alert staff when something conflicts. Human review is especially valuable when evidence is ambiguous, coverage is contested or a decision has serious consequences for a customer. Design the workflow around decision points: define what the system may complete, what triggers referral, who can override it and how the reason is recorded. Measure both speed and decision quality. Faster settlement helps when accuracy and fairness hold; an automated error repeated across thousands of files creates a much larger problem than a single manual error. Feedback from reviewed exceptions should improve the rules and data collection.

Worked example: Assume a covered glass claim has a $500 verified invoice, no deductible and all required fields complete. The system can calculate $500 and route it for automatic payment under an approved rule. A second claim has photographs showing damage inconsistent with the invoice. It is referred to an adjuster to establish what was damaged before coverage and payment are decided.

Watch the distinction: A small claimed amount does not make the facts reliable. Automatic payment is appropriate only when the required evidence and rule conditions are satisfied.

Reference: [8]

20. Put limits around data use

Useful data can still create unacceptable decisions if collected, secured or applied poorly. Define the business purpose, collect only fields needed for it, limit access and retention, and protect data throughout sharing and storage. For a model, examine whether variables have a defensible relationship to the insured exposure, whether decisions can be explained, and whether outcomes differ unfairly across relevant groups under applicable United States state rules. A variable may predict claims while acting as a proxy for a protected characteristic. Test alternatives rather than assuming prediction alone justifies use. Document model inputs, validation results, approval and monitoring so an adverse decision can be traced to evidence. Security and fairness need continuing review because populations, data sources and model behavior change after deployment.

Worked example: An insurer finds that a proposed pricing variable predicts losses but closely tracks a protected characteristic under assumed state rules. It removes or redesigns the variable, retrains the model and compares predictive accuracy and group outcomes on held out data. If the revised model meets the insurer's documented business and fairness criteria, it records the rationale and monitors results after use; the original correlation alone does not justify deployment.

Watch the distinction: Predictive accuracy does not settle whether a variable is lawful, fair or explainable. A model can produce useful forecasts while creating decisions the insurer cannot defend.

Reference: [8]

CPCU 551: Managing Commercial Property Risk

21. Separate building and business property

A commercial property decision starts with who would suffer the loss. A building owner may need building coverage, while a tenant may primarily need coverage for its own stock, equipment and improvements it must repair. Next, identify where each item is located: a scheduled premises, temporary storage site or vehicle can change which coverage applies. Finally, establish the valuation basis. Replacement cost generally considers the cost to replace with comparable property, while actual cash value reflects depreciation; neither automatically equals the sale price. Under a form that insures both categories, apply each category’s limit separately after determining the covered value. Review deductibles, any coinsurance condition and special limits before estimating the final payment. The CPCU 551 course includes building and business personal property decisions.

Worked example: Assume a United States policy covers fire at the listed warehouse, values both losses at the stated amounts, and has separate limits with no deductible or coinsurance adjustment. Fire causes $300,000 of building damage and $80,000 of owned stock damage. The $250,000 building limit caps that payment at $250,000; the $100,000 stock limit permits $80,000. Total payment is $330,000, leaving $50,000 of building damage above its limit.

Watch the distinction: A $100,000 stock limit cannot absorb building damage above the separate building limit. First classify the damaged property, then apply the limit assigned to that category.

Reference: [9]

22. Calculate an income interruption

Business income coverage addresses the financial effect of a covered interruption, not the physical repair bill. Set the covered period of restoration first; revenue lost after that period may require a separate extension. Estimate revenue the firm would have earned during that period from comparable operations, then subtract costs it avoids because work stops. Costs that continue, such as rent or key payroll, remain part of the income exposure. Extra expense is a separate amount spent to resume operations or reduce the interruption, subject to the coverage grant and limits. Compare it with the loss avoided so the calculation does not count the same benefit twice. A seasonal business needs a forecast that reflects the affected months, not a flat annual average. Document the counterfactual sales estimate and any post-loss operating capacity.

Worked example: Assume a covered fire closes a shop during its stated restoration period. Forecast revenue lost is $200,000, and the shop avoids $70,000 of variable purchases and wages. Its income loss is $200,000 minus $70,000, or $130,000. It also spends $20,000 on covered temporary operations that limit the shutdown; assuming no overlap or applicable cap, the combined claim is $150,000.

Watch the distinction: Do not pay every dollar of lost sales as income loss. The shop would have incurred costs to make those sales, so saved costs must be removed.

Reference: [9]

23. Tailor coverage for catastrophe gaps

A large property limit does not mean every cause of loss has that much protection. Work from the actual peril: fire, flood, wind, earthquake and crime may have different grants, exclusions, endorsements and deductibles. Then look for a sublimit, which is a smaller ceiling within the broader policy limit. A percentage catastrophe deductible can also leave a much larger retained amount than a familiar flat deductible, so its stated base matters. Examine whether the relevant limit applies per occurrence, per location or in the aggregate; multiple damaged sites can produce very different results. For a disaster plan, compare the likely severe loss with the applicable limit and the cash available for the deductible. Risk reduction, such as elevating equipment or improving drainage, addresses the remaining physical exposure even when insurance is purchased.

Worked example: Assume a United States property form covers flood at one site but caps it at a hypothetical $100,000 sublimit within a $1 million property limit. A qualifying flood causes $400,000 of covered damage. Before any deductible, the most payable is $100,000, leaving $300,000 above the sublimit. The $1 million headline limit does not increase this flood payment; a deductible could reduce it further.

Watch the distinction: A declarations page showing $1 million is not enough to price a flood loss. Identify the cause, its sublimit and its deductible before using the main limit.

Reference: [9]

24. Protect property away from premises

Property can leave the premises without leaving the business’s financial responsibility. Separate three questions: who owns it, who bears risk of loss under the sales or shipping agreement, and which locations or modes the policy covers. Stock carried by a common carrier may need transit protection; tools that move among jobs may fit an inland marine form; customer property held for repair creates a different interest. A limited off-premises extension can help with an occasional movement but may be too small for routine shipments. Check when transit coverage begins and ends, whether temporary storage counts, and how property in another party’s custody is valued. Map the largest single shipment and the total property exposed at one temporary location before setting limits. Contract terms and insurance terms must be reconciled because they answer different questions.

Worked example: A retailer ships $75,000 of goods. Assume it bears the transit risk, the shipment is covered under an off-premises extension limited to $10,000, and there is no deductible. If all goods are destroyed, that extension pays at most $10,000, leaving $65,000. A transit form with an adequate per-shipment limit would address the gap, provided this route and carrier qualify.

Watch the distinction: Transfer of physical custody does not necessarily transfer financial risk. The shipping contract may leave the seller responsible after the goods enter a carrier’s truck.

Reference: [9]

CPCU 552: Managing Commercial Liability Risk

25. Apply the CGL coverage structure

A commercial general liability policy requires both a liability basis and a fitting coverage grant. Bodily injury and property damage claims are commonly analyzed under one grant; personal and advertising injury, such as specified reputational or publication offenses, is a distinct category. For an occurrence form, ask whether the injurious event happened during the policy period, then test whether the claimant seeks damages because of a covered injury. Read definitions, exclusions and exceptions before limits. An injured employee, damage to the insured’s own work, or an auto accident may call for a different policy or special analysis. Defense obligations and whether defense costs erode a limit depend on the wording. The legal concepts guide helps frame contract interpretation, while the CGL decision still turns on the facts and applicable United States form.

Worked example: Assume a United States occurrence form covers a store’s premises liability, has no applicable exclusion, a $100,000 per-occurrence limit and defense costs outside that limit. A customer trips during the policy period and proves $90,000 of bodily injury damages caused by the store’s negligence. The damages fit the bodily injury grant and limit, leaving $10,000 of that assumed limit unused; defense is considered separately.

Watch the distinction: Do not route every reputational allegation through bodily injury coverage. Personal and advertising injury has its own listed offenses, triggers and exclusions, which must match the allegation.

Reference: [10]; [6]

26. Select commercial auto protection

Vehicle ownership and vehicle use point to different commercial auto exposures. Owned autos belong to the business; hired autos are generally rented or borrowed for business use; nonowned autos can include employees’ personal vehicles used for company work. The coverage symbols or vehicle descriptions determine which category is insured for each coverage. Liability protection addresses the business’s legal responsibility to others, while physical damage protection addresses the covered vehicle itself. An employee’s personal auto policy may respond to a crash, but it does not remove the employer’s possible liability, and the order of payment depends on the policies and facts. Ask who owns the vehicle, who authorized the trip, what task was being done and whether drivers or vehicles fall within the selected coverage. These questions matter especially for delivery work and occasional errands.

Worked example: An employee uses her own car to deliver a customer document and injures another driver. Assume the business can be held liable for work-related driving. Its car is neither owned nor hired; it is a nonowned auto exposure. If the business policy insures only listed owned vehicles, it has no assumed nonowned liability grant for this claim. Adding appropriate nonowned auto liability would address that exposure.

Watch the distinction: An employee’s personal auto coverage is not proof the employer is protected. A claim can name both parties, while the business policy may cover only owned vehicles.

Reference: [10]

27. Separate cyber loss from liability coverage

Cyber incidents can create the insured’s own costs and claims made by others. First-party costs may include forensics, containment, restoration, notification and income lost during a covered system interruption. Third-party liability can arise when customers allege mishandling of their information and seek damages or defense. A cyber policy may use different triggers for each grant: a security failure, privacy event, network interruption or claim first made during the policy period. Waiting periods, sublimits, vendor requirements and consent provisions can change recovery. Traditional liability and property policies may not respond consistently to intangible data loss or breach expenses. Build an event timeline, identify each cost by who incurred it, and match each bucket to its grant and limit. This also shows which risks need prevention, incident response planning or additional insurance.

Worked example: Assume a covered breach produces $40,000 of reasonable forensic costs and a $200,000 customer lawsuit. A hypothetical policy gives $50,000 for approved first-party response costs and $250,000 for covered third-party liability, with no deductible or other claims. The response grant pays $40,000. If the lawsuit establishes covered liability of $200,000, that separate grant pays $200,000. Total assumed payment is $240,000.

Watch the distinction: Do not add breach costs and a lawsuit into one undifferentiated limit. They may have different grants, triggers, sublimits and proof requirements.

Reference: [10]

28. Close workplace and professional gaps

Premises liability concerns injuries tied to ordinary business operations, but workplace and advice-related claims follow different paths. Employee injuries typically raise workers compensation and employers liability questions under the applicable United States state system. Discrimination, harassment or wrongful termination allegations concern employment practices, not physical injury to a visitor. A client’s financial loss from faulty professional advice may require professional liability coverage, often written on a claims-made basis. For each allegation, identify the claimant, the alleged wrongful act and the kind of damages sought. Then test any CGL exclusion, the specialized policy’s trigger and its defense provision. A higher CGL limit increases capacity only for claims the CGL actually covers. Where services and products overlap, examine whether the harm stems from professional judgment or from a defective physical product.

Worked example: A former employee alleges discriminatory firing and seeks $60,000; a client seeks $90,000 for financial harm from faulty consulting advice. Assume the CGL excludes both, while applicable employment practices and professional liability forms each have a $100,000 limit and no deductible. Analyze the claims separately: the first can fit its $100,000 limit, and the second can fit its own $100,000 limit if each form’s trigger and terms are met.

Watch the distinction: Increasing a CGL limit does not remove its exclusions. Coverage for an employment decision or professional error needs an applicable grant in the relevant specialized policy.

Reference: [10]

CPCU 555: Advancing Personal Insurance Products

29. Match home coverage to property interests

A homeowners form divides property by interest and use. The dwelling category concerns the residence and attached structures; other structures generally covers eligible detached structures; personal property concerns belongings; and loss of use can address specified additional living costs after a covered loss. The categories have separate limits and sometimes special restrictions. Classify the damaged item before estimating payment, then test the cause of loss, valuation basis and deductible. A detached garage is not personal property merely because belongings are stored inside it. Business property, valuables or property away from home may have narrower protection or special limits. The CPCU 555 course addresses residential property alongside broader personal insurance needs. In practice, location, occupancy and the household’s ownership interest shape the recommendation.

Worked example: Assume a United States homeowners form covers this tree damage, values repairs at the stated amounts, and has sufficient separate limits with no deductible. The tree causes $30,000 damage to a detached garage and $5,000 damage to stored personal belongings. Other structures coverage pays $30,000; personal property coverage pays $5,000. Total payment is $35,000. Neither item is assigned to the dwelling limit.

Watch the distinction: A detached garage and its contents are not one property category. Mixing them can hide a shortage in the other structures or personal property limit.

Reference: [11]

30. Analyze personal liability exposure

Personal liability analysis begins with an alleged legal obligation to compensate another person. Identify the conduct, injury or damage, causation and possible defenses under the relevant state law. A homeowners liability section may provide defense and indemnity for covered occurrences, subject to exclusions such as certain business or intentional conduct. An umbrella can add a higher layer above required underlying coverage and sometimes provide broader protection, but its attachment and exclusions must be checked separately. Calculate the underlying covered payment first, then the portion above it that falls within the umbrella’s limit. Defense costs can materially affect the result when a form includes them within a limit. A household with frequent guests, rental activity or substantial assets should examine both loss prevention and the size of its liability limits.

Worked example: Assume a covered United States liability judgment is $900,000. The homeowners policy pays its $500,000 limit, and an applicable umbrella attaches immediately above it with a $1 million limit; assume no deductibles or limit-eroding defense costs. The remaining judgment is $900,000 minus $500,000, or $400,000. The umbrella pays $400,000, leaving $600,000 of its own limit unused.

Watch the distinction: An umbrella does not automatically cure every exclusion in the underlying policy. Confirm that the umbrella covers the event and that its attachment requirements are satisfied.

Reference: [11]

31. Choose personal auto coverages

Personal auto coverage is easiest to select by asking whose loss is being paid. Liability coverage addresses injury or property damage the insured legally owes to others. Collision addresses damage to the covered car from impact or overturn; other-than-collision covers specified noncollision causes, often including theft or weather damage. Medical payments or personal injury protection may address occupants’ injuries under the applicable form and state rules. Uninsured or underinsured motorist coverage concerns injury, and sometimes property damage, caused by a driver without adequate liability insurance. Each coverage can have a different deductible or limit. Driver, vehicle and use eligibility still matter. Separate damage to the insured car from damage to another person’s property before applying coverage; one crash can involve several grants rather than a single claim bucket.

Worked example: Assume a driver strikes a pole. The driver’s car has $8,000 of covered collision damage with a $500 deductible, and the pole owner has $3,000 of covered property damage. With sufficient limits and no other adjustment, collision pays $8,000 minus $500, or $7,500; the driver retains $500. Property damage liability pays the pole owner $3,000. Total insurer payment is $10,500.

Watch the distinction: Collision coverage repairs the insured car; it does not pay for the pole. Damage to another owner’s property is evaluated under property damage liability.

Reference: [11]

32. Recognize wider household needs

Household planning connects risks without treating insurance products as interchangeable. Start with property that ordinary homeowners terms may restrict, such as valuable collections or business equipment, and identify whether an endorsement or separate form is needed. Next, estimate the financial impact of death or disability: debts, future income needs, dependants and available assets determine the gap more than a single salary multiple. Health coverage addresses medical expense exposure, while savings provide liquidity for deductibles, temporary income disruption and planned expenses. Life coverage can fund survivors’ needs but cannot repair a house or pay a current medical bill for a living insured. Use one consistent time basis when estimating future needs and avoid counting an asset both as a debt offset and as a source for another goal. Revisit the estimate when family obligations change.

Worked example: A family estimates $400,000 of debt and $200,000 of additional income-replacement need, after making its own time-horizon assumptions. Total stated need is $600,000. Existing life cover is $150,000, so the preliminary gap is $600,000 minus $150,000, or $450,000. This is a hypothetical planning estimate; other assets, survivor income, taxes and future expenses could change it.

Watch the distinction: A $450,000 life-cover gap does not describe the family’s health or property insurance need. Those losses occur under different conditions and require separate estimates.

Reference: [11]

CPCU 556: Building a Competitive Edge in Personal Lines

33. Choose a personal lines market

Distribution strategy depends on the customers an insurer wants to serve and the lifetime economics of reaching them. A low acquisition cost helps only if the channel brings suitable risks, customers understand the product and policies stay in force long enough to recover acquisition spending. Online quoting suits standardized submissions; an adviser may add value when households need coverage explanations or unusual exposures assessed. Compare expected premium contribution after claims and service costs, retention, acquisition cost and any channel-specific expense. Then check whether underwriting results differ because channels attract different risks. A channel that produces rapid growth can weaken a portfolio if claims or cancellations rise. Product design matters too: clear choices and service support can reduce misunderstanding and complaints. Use comparable customer cohorts when judging channel performance, since a raw cost comparison confuses price with value.

Worked example: Assume each retained policy contributes $100 per year after expected claims and service costs. Online acquisition costs $80 and customers stay two years on average, giving $200 minus $80, or $120 expected contribution. Agent acquisition costs $140 but customers stay three years, giving $300 minus $140, or $160. Under these simplified assumptions, the agent channel yields $40 more per acquired policy despite its higher acquisition cost.

Watch the distinction: Choosing the cheaper acquisition channel from its first invoice ignores retention and claim quality. Compare expected contribution over the customer relationship using comparable groups.

Reference: [12]

34. Evaluate personal auto risk

Personal auto underwriting combines driving history, vehicle characteristics, territory, mileage and use. Each describes a different part of expected loss: how often the vehicle is exposed, the likely severity of an incident, and the driver’s observed record. Verify the stated use before applying a rating or eligibility rule. Delivery use may change driving patterns and may require a different product from ordinary personal use, even when the vehicle is identical. Annual mileage is informative but incomplete; driving time, route and purpose also matter. Investigate contradictions such as low reported mileage paired with daily commercial deliveries, and use evidence rather than treating one signal as proof of misconduct. A defensible decision records the verified facts, the applicable filed or approved underwriting rule where relevant, and the reason a quote, referral or decline follows.

Worked example: Two applicants own the same car. One reports 4,000 annual miles for personal errands; the other reports 25,000 annual miles including deliveries. The second drives 25,000 divided by 4,000, or 6.25 times as many miles. That does not prove losses will be 6.25 times higher, but it warrants verification of mileage and delivery use before selecting an eligible policy and applying its rules.

Watch the distinction: Equal vehicles do not imply equal auto exposure. Mileage and delivery work can change frequency and eligibility, while the mileage ratio alone is not a rate multiplier.

Reference: [12]

35. Assess residential and special exposures

A residential risk has several dimensions that should be assessed separately. Roof age and condition inform water and wind vulnerability; location informs catastrophe and theft exposure; occupancy affects how quickly leaks or intrusions are discovered. Unusual use, including home business storage or short-term rental activity, can add both property and liability exposures. A favorable feature does not cancel an unrelated hazard. Verify the length and reason for vacancy, whether someone checks the house, whether utilities remain on and where business inventory is stored. Then compare those facts with the insurer’s eligibility rules and the coverage available for each property type. Possible responses include protective measures, different terms or a specialized product. The goal is to identify the mechanism of loss and the remaining gap, not to reduce the household to a single good-or-bad score.

Worked example: A home has a new roof, will be vacant for six months and holds $20,000 of business inventory. The roof lowers concern about roof deterioration, but it does not prevent an undetected leak during vacancy. Assume the homeowners form covers only $5,000 of that business property: $20,000 minus $5,000 leaves a $15,000 potential property gap. Review vacancy terms and arrange suitable business property protection separately.

Watch the distinction: A new roof does not make a vacant house an ordinary occupied risk. It addresses one cause of loss while vacancy changes detection and response.

Reference: [12]

36. Steer a personal lines portfolio

Portfolio results should be measured on consistent periods and exposures before changing price or terms. A combined ratio adds incurred losses and underwriting expenses, then divides by earned premium. Above 100% indicates a simplified underwriting loss before investment income; it does not reveal the cause. Split the result by segment, accident period, geography, claim type, acquisition channel and expense source. Ask whether a few large claims, rising frequency, inflation, inadequate rates or distribution costs drive the result. Check reserve development and data maturity before treating one period as a stable trend. Responses differ: prevention suits avoidable losses, pricing suits credible rate inadequacy, and product or eligibility changes suit risks outside the intended portfolio. Evaluate the likely effect on customers, retention and applicable state rate rules before recommending a change.

Worked example: A personal lines segment earns $5 million premium, incurs $4 million in losses and $1.2 million in underwriting expenses. Its loss ratio is $4 million divided by $5 million, or 80%; its expense ratio is 24%. The combined ratio is 104%, meaning a simplified $200,000 underwriting loss. Diagnose whether claims or expenses are unusually high before selecting a remedy.

Watch the distinction: A 104% combined ratio does not automatically justify a 4% rate increase. Mix, claim development and customer response can make that arithmetic misleading.

Reference: [12]

ACRM 401: Effectively Managing Cyber Risk

37. Map a cyber loss pathway

A cyber loss pathway connects an actor or error to an entry point, a vulnerable asset and a consequence. The actor might be a criminal using stolen credentials, an insider misusing access or an employee making a mistake. A vulnerability is the condition that lets the event progress, such as excessive privileges or weak authentication. The affected asset might be customer data, operating software or payment capability. Consequences can include direct response expense, interrupted operations, third-party claims and reputational harm. Trace the steps in order because a control works at a particular point: strong authentication can impede entry, limited privileges can constrain spread, and tested backups can shorten downtime. The ACRM 401 course includes vulnerability assessment, data protection and loss mitigation. A useful map names both the initial compromise and the business harm it can produce.

Worked example: An attacker uses a stolen password to enter a billing system. Weak authentication permits entry; broad account privileges permit access to customer records and billing functions. The firm faces data exposure and a separate period of downtime. Require stronger authentication, narrow privileges and monitor unusual access to interrupt different steps. Backups may help restore billing, but they cannot retrieve records already copied by the attacker.

Watch the distinction: Calling the stolen password the entire loss misses the chain. The credential is an entry route; data disclosure and operational interruption are distinct consequences.

Reference: [13]

38. Prioritize cyber vulnerabilities

Prioritize cyber weaknesses by credible loss pathway, not by the number of findings on a scan. Estimate how likely exploitation is, what assets the weakness reaches, how long interruption could last and how well current controls would detect or limit it. A publicly exposed administrator account that reaches production systems can dominate an outdated brochure because its possible operational impact is much greater. Consider dependencies: a single identity system may control several critical applications. Then compare feasible controls, their cost and the residual risk after implementation. Expected loss can help rank options, but uncertain probabilities and severe tail outcomes require judgment beyond one average. Record assumptions so a change in exposure, threat activity or control effectiveness can change the ranking. Confirm that fixing the highest-ranked weakness actually removes the exploitable path rather than merely closing a visible ticket.

Worked example: Assume an exposed administrator account could cause one day of interruption costing $2 million, with a hypothetical 2% annual chance if left open. Its simplified expected annual loss is $40,000. A stale public brochure has no system access and only a $1,000 plausible correction cost. Secure the account first with access restrictions and monitoring; the $40,000 estimate is a ranking aid, not a guaranteed saving.

Watch the distinction: A long list of low-impact findings is not necessarily more urgent than one critical access path. Rank by plausible consequence and exploitability, then verify control effectiveness.

Reference: [13]

39. Protect data through its life cycle

Data risk changes as information is collected, used, stored, shared and disposed of. Collect only information needed for a defined purpose; excess records increase what a breach can expose. Classify sensitive data so access can be limited by role, reviewed when jobs change and logged for investigation. Protect stored and transmitted data, but remember that encryption does not help if an attacker operates through an authorized account with broad access. Retention rules determine when records should be deleted or securely destroyed, including copies held by vendors. Tested backups protect availability after deletion, ransomware or system failure; they do not reverse information already disclosed. Human controls such as training and approval checks complement technical controls. Map each life-cycle stage to a likely failure and a control, then test whether the control works in normal operations.

Worked example: A firm holds 50,000 former customers’ identity files indefinitely, although only 5,000 are still needed for a stated purpose. Under a hypothetical approved retention plan, secure disposal of the other 45,000 reduces retained files by 90%. Restrict access to the remaining 5,000 and test backups. If records were already copied, later deletion and backup restoration would not undo that disclosure.

Watch the distinction: Backups solve availability problems; they do not erase an attacker’s copy of disclosed data. Retention limits and access controls address that different harm.

Reference: [13]

40. Plan response and legal review

Incident response follows parallel operational and legal tracks. First contain ongoing access without destroying evidence: disable compromised credentials, isolate affected systems as appropriate and preserve logs, images and decision times. Establish which data and systems were reached, which people may be affected and whether access became actual acquisition, where that distinction matters under applicable law. Restore from verified clean systems while monitoring for renewed intrusion. In a United States incident, notification duties may differ by affected residents’ states, data type, contractual promises and sector rules, so determine the jurisdictions before setting a deadline or message. Coordinate legal, security, operations and insurance contacts so facts stay consistent. Record what is known, uncertain and still being investigated; premature conclusions can create inaccurate notices, while needless delay can worsen harm. After recovery, test the control changes against the original entry route.

Worked example: A United States firm discovers unauthorized billing-system access involving residents of two states. It disables the account, preserves access logs and identifies the affected records while restoring billing from a verified clean environment. Its response team separately evaluates each state’s applicable notification rule and any contracts before setting deadlines. It documents which residents and data are confirmed, then sends notices required by the resulting legal assessment.

Watch the distinction: One incident does not imply one universal United States notification deadline. Determine the affected jurisdictions and data first, while preserving evidence and containing access promptly.

Reference: [13]

AIC 300: Claims in an Evolving World

41. Build a defensible claim file

A defensible claim file shows what was reported, what was observed, when each event occurred, and why the evidence supports the conclusion. Start with the insured’s account, policy details, reported damage and relevant dates. Interview witnesses separately so one account does not shape another. Preserve photographs, inspection notes, repair records and communications with their sources and dates. Build a chronology, then mark disputed points for investigation. An allegation that a pipe burst is different from a photograph showing wet flooring; neither alone establishes when the water escaped. Earlier maintenance reports may affect the cause, the amount of damage attributable to each period, or both. Record reasoning as the file develops, including evidence that challenges the initial theory. This lets another claims professional understand the decision and explain it fairly to the claimant.

Worked example: A claimant reports a Monday pipe burst. A Friday maintenance ticket describes a small leak in the same wall. The adjuster records both dates, obtains the ticket’s author’s account, photographs the pipe and flooring, and asks a qualified inspector to distinguish older staining from fresh damage. The Friday report raises a timing question; it does not establish that all Monday damage occurred earlier.

Watch the distinction: A prior leak report proves the entire claim is old damage. It proves only that a leak was reported earlier; the file still needs evidence about the later event and resulting damage.

Reference: [14]

42. Make a coverage determination

A coverage decision should follow a consistent sequence. Identify the insured, property, loss date and applicable form. Test whether the facts meet the insuring agreement, then apply definitions, exclusions and any exceptions to exclusions. Check conditions such as notice and cooperation, while assessing their effect under the assumed United States policy and applicable state law. Separate physical causes and damage periods when the evidence supports allocation. An exclusion for long-term seepage does not automatically remove distinct damage caused by a later sudden release. Conversely, a covered pipe break does not convert preexisting deterioration into covered damage. Use inspection findings and a reasoned estimate to assign amounts, then apply valuation, limits and deductible to the eligible portion. Explain both the payable and unpaid portions in terms the claimant can trace to facts and policy wording.

Worked example: Assume a United States property form covers sudden pipe damage, excludes long-term seepage, and has a $2,000 deductible with sufficient limit. Inspection supports $12,000 of new damage from a sudden break and $8,000 of older seepage damage. The eligible loss is $12,000; after the deductible, payment is $10,000. The $8,000 remains outside this assumed coverage.

Watch the distinction: Once a covered event occurs, every item found during inspection is covered. Coverage follows each item’s supported cause and timing, so older excluded damage must be evaluated separately.

Reference: [14]

43. Settle on supported damages

A fair liability settlement begins with evidence of fault and damages, not a single negotiating target. Confirm the claimant’s injuries or property loss with records, invoices and, where needed, expert analysis. Examine defenses, comparative fault and the likely effect of disputed evidence under the assumed state law. Estimate the range of possible trial outcomes, then add expected defense expense, delay and uncertainty. Expected value is a planning measure: probability of liability multiplied by the supported award, plus expected defense cost. It can help set negotiating authority, but an individual trial will produce an actual result rather than an average. Consider whether a proposed settlement releases the right claims and whether payment terms are clear. Document why the amount reasonably addresses both the claimant’s demonstrated harm and the insurer’s litigation exposure.

Worked example: Assume a United States liability claim has $30,000 in supported damages, a 60% assessed chance of liability, and $8,000 in expected defense expense if litigated. The simplified expected litigation cost is 0.60 × $30,000 + $8,000 = $26,000. That figure informs negotiation; stronger evidence, a different likely award or a release covering other claims could change the appropriate offer.

Watch the distinction: The $26,000 expected cost is the settlement owed. It is a decision estimate under stated assumptions, while fair settlement depends on supported damages, legal exposure and negotiated terms.

Reference: [14]

44. Use fraud indicators responsibly

Fraud indicators identify claims worth closer examination; they are not findings of dishonesty. An unusually high estimate might reflect inflated pricing, but it might also reflect concealed damage, difficult access or regional labor costs. Compare like repairs and document the benchmark’s scope before asking the claimant or contractor to explain the difference. Seek records and an independent inspection proportionate to the concern. Preserve ordinary claim service while investigating, avoid accusations unsupported by evidence, and follow the jurisdiction’s claim handling rules. Separate an inaccurate statement from an intentional material misrepresentation: memory errors and innocent mistakes can produce inconsistencies. The final file should show the indicator, steps taken, evidence obtained and reasoning for payment, further investigation or a supported denial. That discipline protects legitimate claimants and makes referrals based on actual findings rather than a model score or suspicion.

Worked example: A model flags a $40,000 repair estimate against a $20,000 local benchmark, a 2:1 ratio. The benchmark covers visible finishes only. An engineer documents concealed structural damage costing another $20,000, bringing supported repair cost to $40,000. The adjuster records the scope difference and handles the otherwise covered claim on that evidence; the flag supplies no independent ground for denial.

Watch the distinction: A fraud score or high price is proof of fraud. It is an investigation lead; the decision requires evidence about the loss, the estimate and any allegedly false statement.

Reference: [14]

AIDA 401: Using Data Analytics to Strengthen the Insurance Value Chain

45. Choose trustworthy analytical inputs

Analytical inputs must describe the exposure and outcome the decision actually concerns. Structured fields, such as claim date and paid amount, are easy to aggregate but may carry inconsistent definitions. Unstructured adjuster notes, images and correspondence can explain causes, though they need careful coding and quality checks. Confirm the observation period, source system, missing fields, duplicates and changes in claim or policy definitions. Use a denominator that matches the question: claim count measures workload, while claims per exposed policy measures frequency. A chart with counts can rise simply because the book grew. Compare groups with similar coverage, geography and time at risk before attributing a change to underwriting or prevention. The AIDA 401 course addresses data interpretation across insurer operations; a useful analysis still depends on clear definitions and credible records.

Worked example: A dashboard shows 100 claims among 1,000 policies in one year and 120 among 1,500 in the next. Assuming each policy represents the same full-year exposure, frequency changes from 100 ÷ 1,000 = 10% to 120 ÷ 1,500 = 8%. Claim workload rose by 20, but observed frequency fell by 2 percentage points. Check exposure duration before relying on that comparison.

Watch the distinction: More claims necessarily mean higher risk per policy. A growing portfolio can generate more claims while frequency falls, so compare claims with a consistent exposure denominator.

Reference: [15]

46. Find the claim process bottleneck

Process data become useful when every stage has a clear start and end event. Define intake, investigation, approval and payment consistently, then compare both typical duration and the slow tail. A long investigation stage may reflect waiting for documents, expert scheduling, complex coverage questions or staff capacity; the duration alone does not name the cause. Segment claims by type and complexity so a large loss is not treated as equivalent to a simple glass claim. Examine queue time separately from active work time, because adding staff will have different effects from improving information collected at first notice. When changing the process, track accuracy, reopened claims and claimant outcomes alongside speed. Faster closure that later requires correction may increase total work and harm service. Use the bottleneck evidence to choose a specific intervention and measure its effect.

Worked example: Suppose a simplified claim journey has median stage times of 2 days for intake, 14 for investigation, 3 for approval and 1 for payment. The stated total is 2 + 14 + 3 + 1 = 20 days, so investigation deserves first review. If 9 of its 14 days are spent waiting for repair documents, earlier document requests are a plausible intervention; check results on comparable claims.

Watch the distinction: The longest stage automatically needs more adjusters. Its delay could arise from missing documents, external experts or handoffs; diagnose queue time and work time first.

Reference: [15]

47. Test an underwriting or marketing model

A predictive model ranks prospects or risks; evaluating it requires outcomes from the population where it will be used. Define the target, such as purchase within a set period or claim frequency over an exposure year. Use separate training and evaluation data, and check calibration or error by relevant segment. A marketing campaign adds another problem: selected customers may already be more likely to buy. Comparing them with an unselected group gives an observed difference, not necessarily the campaign’s causal effect. A randomized holdout, where feasible and appropriate, estimates incremental purchases more credibly. For underwriting, compare predicted risk with later losses while checking for changing exposure mix and unacceptable proxy effects. A model that separates groups well may still produce poor decisions if its predictions are miscalibrated or its data do not match the new market.

Worked example: A campaign reaches 1,000 selected customers and 200 buy, giving 200 ÷ 1,000 = 20% conversion. A comparable group shows 15%, so observed lift is 5 percentage points, or 50 additional purchases per 1,000 customers at that rate. Because selection may favor likely buyers, those 50 cannot be credited to the campaign without a credible control design.

Watch the distinction: A higher conversion rate proves the model or campaign caused more sales. Selection can explain the gap; a suitable holdout is needed to estimate incremental effect.

Reference: [15]

48. Measure prevention value

Prevention value is the expected reduction in loss cost less the full cost of the intervention. Begin with a defined exposure base and time horizon. Estimate claim frequency and severity with and without the technology, then include device purchase, installation, monitoring, maintenance and response costs. Separate losses the device can plausibly prevent from those it merely detects after damage begins. Compare similar properties or use a credible trial, because customers who adopt sensors may also maintain their buildings better. Check whether the insurer, insured or both receive the savings; that affects pricing, incentives and adoption. Expected net benefit is a portfolio planning figure, not a guaranteed annual cash saving. A rare severe loss can make any single year look very different from the modeled average. Review performance after deployment and revise assumptions when measured outcomes diverge.

Worked example: For a hypothetical group of buildings, annual expected water claims fall from $180,000 without sensors to $110,000 with them, a $70,000 reduction. Annualized device, installation and monitoring costs total $50,000. Expected net benefit is $70,000 − $50,000 = $20,000 per year. The decision is favorable only if comparable exposure data support that reduction and the costs include continuing operation.

Watch the distinction: Subtracting device price from last year’s claims proves savings. Last year may be unusual, and annual costs or exposure differences may be missing from the comparison.

Reference: [15]

ARe 321: Exploring Reinsurance Products and Transactions

49. Use proportional reinsurance

Proportional reinsurance divides premium and covered losses by an agreed share. Under quota share, the percentage applies across covered business: a 40% cession leaves the insurer with 60% of each covered risk and loss, subject to treaty terms. Surplus share instead starts with the insurer’s monetary retention, often expressed as one line. The reinsurer takes the portion above that line up to the treaty’s available line capacity; the ceded percentage therefore varies by risk. Calculate the proportion at inception from the insured amount, then apply that proportion to covered losses on that risk. A risk above combined insurer retention and treaty capacity needs another placement or a lower accepted amount. Commissions, treaty exclusions and limits can affect the financial result, so a simple loss split does not establish the final net profit.

Worked example: On a $10 million covered risk, a hypothetical 40% quota share cedes $4 million of exposure and $400,000 of a $1 million covered loss; the insurer retains $600,000. Separately, suppose a $6 million risk has a $2 million insurer retention and adequate surplus treaty capacity. The ceded share is $4 million ÷ $6 million = two-thirds, so a $900,000 covered loss splits $600,000 to the reinsurer and $300,000 to the insurer.

Watch the distinction: Surplus share always cedes a fixed percentage like quota share. Its proportion depends on each risk’s insured amount relative to the insurer’s retention and available treaty lines.

Reference: [16]

50. Protect individual large losses

Excess of loss reinsurance responds after a defined loss unit reaches an attachment amount. A property per risk layer usually tests the covered loss from one insured risk against its retention, then pays the excess up to the layer limit. Casualty excess can use a different loss unit, such as one occurrence or one claim, and its treatment of defense costs can materially alter attachment and recovery. Identify that unit before calculating. For a layer described as $3 million excess of $1 million, recovery is the lesser of $3 million and the covered loss above $1 million, never below zero. The insurer bears the first $1 million and amounts above the layer’s top unless other protection applies. Multiple damaged locations, related claims and reinstatements require attention to contract definitions. The ARe 321 course lists both property per risk and casualty excess topics.

Worked example: Assume one building’s covered loss is $2.5 million and its property per risk treaty provides $3 million excess of $1 million, with no relevant exclusions. The amount above attachment is $2.5 million − $1 million = $1.5 million, below the $3 million layer limit. The reinsurer pays $1.5 million; the insurer retains $1 million. A second risk would be tested separately under this assumption.

Watch the distinction: Every large claim under the policy is pooled before applying a per risk attachment. The treaty’s defined loss unit controls which losses combine and where the layer begins.

Reference: [16]

51. Protect catastrophe and annual totals

Catastrophe excess and aggregate excess protect against different patterns. An occurrence catastrophe layer groups covered losses arising from a defined event and applies its attachment and limit to that event. The event definition, geography, covered causes and hours provisions determine which losses can be combined. Aggregate excess instead looks at cumulative qualifying losses during the contract period and begins paying after an annual attachment is exhausted, subject to its limit. Keep both calculations separate: a moderate series of unrelated losses might reach an aggregate layer without any one event reaching catastrophe attachment. Conversely, one severe event may pierce a catastrophe layer while annual aggregate terms still produce a different recovery. Check whether losses are measured gross or net of other reinsurance, as specified by the contracts, to avoid double counting. These structures stabilize different parts of an insurer’s loss distribution.

Worked example: A hypothetical catastrophe layer covers $10 million excess of $5 million per qualifying event. One event produces $12 million of covered losses, so recovery is $12 million − $5 million = $7 million, within the $10 million limit. Separately, assume an annual aggregate layer covers $4 million excess of $9 million in cumulative qualifying losses. If annual losses total $11 million, it pays $2 million under those separate terms.

Watch the distinction: An aggregate treaty pays whenever one catastrophe crosses the aggregate attachment. Aggregate attachment is tested against qualifying cumulative losses over its stated period, with any other reinsurance treatment specified by contract.

Reference: [16]

52. Read the reinsurance transaction

A projected recovery is an asset only to the extent the transaction supports it. Read the treaty’s covered business, effective period, territory and cause of loss before calculating any layer payment. Then test the defined loss unit, attachment, limit, exclusions and whether other recoveries reduce the subject loss. Reporting and claims cooperation provisions may affect how the cedent presents and develops a claim. A valid ceded amount also carries counterparty credit risk: a reinsurer’s inability to pay can leave the cedent financing the gross claim. Claims, finance and reinsurance staff should reconcile the ceded estimate with the actual contract and supporting loss records. Where interpretation is disputed, document competing readings and pursue the contract’s resolution route. Do not book an assumed recovery merely because the event is large or a broker illustration suggested protection.

Worked example: A cedent models a $6 million recovery from a severe event, but the signed treaty excludes that cause. If no exception or other treaty applies, the supported recovery under this treaty is $0, not $6 million. The cedent must fund the extra $6 million net exposure compared with its model, correct its estimate, and raise the wording issue with the reinsurer using the signed terms.

Watch the distinction: A large loss that reaches the layer must be reimbursed. Attachment matters only after the loss satisfies the treaty’s scope, period, loss definition and exclusions.

Reference: [16]

AU 300: Strengthening Commercial Underwriting Decision-Making

53. Weigh conflicting underwriting evidence

Commercial underwriting begins by deciding which facts describe the actual operation. An applicant’s label may be convenient but too broad to capture heat, chemicals, products or public access. Compare the application with site reports, loss runs, financial records and credible third party observations. Give more weight to verified, recent, relevant facts while checking whether apparent conflicts arise from different dates or locations. Ask targeted questions that could change classification, terms or price: what process occurs, how often, at which premises and under what controls? A missing answer is an information gap, not proof that the worst case applies. Record the material uncertainty and choose whether to quote conditionally, inspect further or defer a decision. The commercial property risk guide can help frame how building use and protection affect the property exposure.

Worked example: An application describes a plant as light assembly, but a recent site report photographs spray finishing. The underwriter confirms the report is for the proposed location, then requests the finishing materials, booth arrangement, ventilation and fire protection details. Until verified, a quote based solely on light assembly understates the potential fire exposure. The underwriter revises classification or terms after assessing the process and controls.

Watch the distinction: The application label governs because the applicant supplied it. Verified operational evidence can reveal a different exposure; the underwriter should resolve material conflicts before relying on the label.

Reference: [17]

54. Map the commercial exposure

Account suitability depends on the whole loss pathway. For property, assess construction, occupancy, protection, neighboring hazards and interruption potential. For liability, consider visitors, products, completed work, contracts and potential injury severity. Operations connect these categories: a production failure can damage stock, halt sales and lead to customer claims. Controls should match the exposure they address. Sprinklers can reduce fire severity, but they do little to prevent mislabeled ingredients from reaching consumers. Examine quality checks, traceability, contingency planning and management’s response to prior incidents. Then compare the account with the insurer’s appetite, available coverage and capacity. A favorable feature in one category does not cancel a serious gap in another. The underwriting decision should state which hazards are accepted, which need mitigation and which require different coverage or capacity.

Worked example: A hypothetical food processor operates in sprinklered buildings and has good fire inspection records, but lacks a product recall plan. The sprinklers support a more favorable property assessment. For product exposure, the underwriter asks how batches are traced, contaminated goods are withdrawn and customers are notified. If those controls remain absent, liability terms or account acceptance may change despite the favorable building protection.

Watch the distinction: Strong property protection makes the entire account low risk. Product contamination can create liability and interruption losses that sprinklers do not address.

Reference: [17]

55. Read the applicant's finances

Financial review asks whether the applicant can operate, invest in controls and fund retained losses. The current ratio compares current assets with current liabilities, but it does not show how quickly receivables or inventory turn into cash. Examine cash, receivable quality, debt maturities, operating cash flow and available credit. Leverage can magnify strain when losses coincide with weak sales or a required capital project. Compare a proposed deductible with realistic cash needs at the time of loss, including repair deposits and ongoing payroll. A profitable firm can still face a liquidity shortage; a high cash balance may be committed elsewhere. Use trends and industry context rather than a single ratio. Financial strength supports the ability to retain loss, while operational controls determine much of the likelihood and severity of the loss itself.

Worked example: A firm has $3 million in current assets and $2 million in current liabilities, so its current ratio is $3 million ÷ $2 million = 1.5. Suppose only $200,000 is cash and $1.4 million is slow moving inventory. A proposed $1 million deductible could exceed immediately available funds. Review cash flow and committed credit before accepting that retention as financially workable.

Watch the distinction: A 1.5 current ratio proves the firm can pay a $1 million deductible. The ratio includes assets that may be unavailable when a claim requires cash.

Reference: [17]

56. Support an underwriting decision

A defendable underwriting recommendation joins exposure evidence, control quality, finances and available terms. First identify the causes and sizes of plausible losses. Compare them with the insurer’s appetite, pricing assumptions and capacity. Decide whether the risk is acceptable as presented, acceptable after modification, or unsuitable. A modification must be specific enough to verify: what improvement, at which location, by when, and what happens if it is not completed? Price the residual exposure after credible improvements, rather than assuming a promised control already works. A deductible can align incentives and reduce small loss cost, but it cannot remove severe loss potential and must fit the insured’s liquidity. Explain the recommendation in a short chain of evidence, decision and conditions. That makes review possible and gives the producer a clear account of what could change the offer.

Worked example: A warehouse has strong cash flow but repeated fires in an unsprinklered storage area. The underwriter offers a hypothetical conditional quote requiring verified sprinkler protection before coverage begins, with a price reflecting the remaining protected storage exposure. If installation is not verified, the offer does not proceed on those terms. Strong finances support a possible deductible, but they do not reduce ignition or fire spread.

Watch the distinction: A wealthy applicant offsets poor fire protection. Financial resources can fund retention and repairs; they do not prevent a fire or make an unpriced severe exposure acceptable.

Reference: [17]

Ethics 311: Ethical Decision Making in Risk and Insurance

57. Use an ethical decision sequence

Ethical decisions need a repeatable sequence when commercial pressure competes with a duty. State the facts and identify what remains uncertain. List affected parties, including the customer, insurer, colleagues and anyone relying on the transaction. Identify legal, contractual and professional duties, then compare realistic actions and foreseeable harms. Ask whether the choice would remain defensible if the full facts were disclosed. Correcting a known material application error protects accurate underwriting and the customer’s ability to rely on the issued coverage. Delaying correction to preserve a sale can create later disputes and erode trust. Escalate when authority or duties are unclear, and record both the action and the explanation given. The Ethics 311 course covers frameworks and workplace dilemmas; the useful habit is to apply one before the decision becomes harder to reverse.

Worked example: Before binding, a producer notices that an application lists a warehouse as sprinklered although inspection confirms it is not. The producer corrects the application, sends the new fact to underwriting, explains any revised quote to the customer and records the exchange. Binding at the lower premium would leave the insurer pricing the wrong risk and could leave the customer facing a coverage dispute.

Watch the distinction: An error can be fixed after the sale if no loss occurs. A known material fact belongs in the underwriting decision before binding, regardless of whether a claim later arises.

Reference: [3]

58. Manage a conflict of interest

A conflict of interest exists when compensation, relationships or personal benefit could influence professional advice. It does not automatically mean every recommendation is wrong, but it calls for a decision process the client can examine. Define the client’s needs, then compare suitable options on coverage scope, exclusions, limits, deductibles, service and total cost. Keep compensation separate from the fit analysis. Disclose material conflicts as applicable under the relevant United States state rules and organizational standards; disclosure alone does not make unsuitable advice acceptable. If the conflict is substantial, obtain independent review or transfer the decision to someone without the incentive. Record the options considered and why the recommended one best meets the client’s stated priorities. This protects trust when a producer’s financial interest and the customer’s interest point in different directions.

Worked example: Two hypothetical policies cover a client’s premises exposure adequately. Policy A costs $8,000 and pays the producer $1,200; Policy B costs $7,500, has comparable relevant terms and pays $750. If service and insurer quality are also comparable, B saves the client $500 and is the better fit despite $450 less producer compensation. The producer documents the comparison and handles required conflict disclosure.

Watch the distinction: Disclosing a higher commission makes the higher priced policy the right choice. Disclosure addresses transparency; the recommendation still needs a client centered coverage and cost rationale.

Reference: [3]

59. Protect confidential information

Confidential claim information should flow only to people who need it for an authorized task. Identify the specific decision, the minimum data it requires and the approved channel before sharing. Medical records can contain unrelated diagnoses and sensitive details; a payment reviewer may need a verified treatment amount without needing the entire file. Check identity and role, use access controls, and leave an audit trail of what was shared and why. Urgency may justify prompt review, but it does not broaden the recipient’s authority. If an outside expert needs records, limit the scope to the question assigned and follow the applicable consent, contract and United States privacy requirements. Secure retention and disposal matter after the claim closes because unnecessary copies increase exposure. Good service and confidentiality work together when information is routed precisely.

Worked example: A payment colleague asks for a claimant’s complete medical file to release a $4,000 approved reimbursement. The claims professional confirms the colleague is authorized to process payment and sends the approved amount and necessary billing reference through the secure claims system. Unrelated medical history stays restricted. If the payment requires another fact, the colleague can request that specific item with a stated purpose.

Watch the distinction: A coworker involved in the claim can see every record. Access depends on the coworker’s authorized task and the information needed to complete it, not general involvement.

Reference: [3]

60. Escalate a harmful practice

When a routine practice appears harmful, identify the instruction precisely and preserve the records needed to assess it. Distinguish a disputed business judgment from a directive that bypasses material evidence or fair review. Note affected claims, dates, the evidence left unread and any immediate customer harm. Raise the concern through an appropriate supervisory, compliance or ethics channel under organizational procedures, using a route outside the implicated supervisor when necessary. Avoid altering files or making unsupported accusations. Seek a timely review of pending claims so escalation does more than create a report. Document responses and follow up if the practice continues. In the United States, claim handling duties and reporting protections can vary by state, so the applicable rules matter. The practical aim is a supported claim determination for each customer and a corrected process.

Worked example: A supervisor tells adjusters to close ten disputed claims without reading newly submitted repair reports. An adjuster preserves the reports and closure instruction, identifies the ten files, and raises the issue with an authorized compliance channel outside that supervisor’s decision path. The requested remedy is review of the new evidence before final determinations. The adjuster records the response and checks that affected claimants receive corrected decisions where warranted.

Watch the distinction: Following a supervisor’s routine instruction removes personal responsibility. A claims professional should escalate a practice that skips material evidence and help secure fair review for affected customers.

Reference: [3]

Official references and further reading

All study guides

Choose your exam version

CPCU500

CPCU520

CPCU530

CPCU540

CPCU550

CPCU551

CPCU552

CPCU555

CPCU556

ACRM401

AIC300

AIDA401

ARE321

AU300

CPCU 500: Becoming a Leader in Risk Management and Insurance

1. Classify the risk before choosing a remedy

2. Measure frequency and severity together

3. Build a layered risk treatment

4. Defend a strategic recommendation

CPCU 520: Meeting Challenges Across Insurance Operations

5. Trace the insurance value chain

6. Connect claims work to the insurer's promise

7. Coordinate underwriting and distribution

8. Link pricing to capacity

CPCU 530: Applying Legal Concepts to Insurance

9. Read the insurance contract in sequence

10. Identify the property interest

11. Work through a tort claim

12. Assess authority and dispute routes

CPCU 540: Contributing to Insurer Financial Performance

13. Read insurer accounts with their purpose in mind

14. Reconcile premium and claim measures

15. Diagnose underwriting ratios

16. Balance investments and capital

CPCU 550: Maximizing Value with Data and Technology

17. Test whether data answer the question

18. Interpret a risk model

19. Place technology where it helps

20. Put limits around data use

CPCU 551: Managing Commercial Property Risk

21. Separate building and business property

22. Calculate an income interruption

23. Tailor coverage for catastrophe gaps

24. Protect property away from premises

CPCU 552: Managing Commercial Liability Risk

25. Apply the CGL coverage structure

26. Select commercial auto protection

27. Separate cyber loss from liability coverage

28. Close workplace and professional gaps

CPCU 555: Advancing Personal Insurance Products

29. Match home coverage to property interests

30. Analyze personal liability exposure

31. Choose personal auto coverages

32. Recognize wider household needs

CPCU 556: Building a Competitive Edge in Personal Lines

33. Choose a personal lines market

34. Evaluate personal auto risk

35. Assess residential and special exposures

36. Steer a personal lines portfolio

ACRM 401: Effectively Managing Cyber Risk

37. Map a cyber loss pathway

38. Prioritize cyber vulnerabilities

39. Protect data through its life cycle

40. Plan response and legal review

AIC 300: Claims in an Evolving World

41. Build a defensible claim file

42. Make a coverage determination

43. Settle on supported damages

44. Use fraud indicators responsibly

AIDA 401: Using Data Analytics to Strengthen the Insurance Value Chain

45. Choose trustworthy analytical inputs

46. Find the claim process bottleneck

47. Test an underwriting or marketing model

48. Measure prevention value

ARe 321: Exploring Reinsurance Products and Transactions

49. Use proportional reinsurance

50. Protect individual large losses

51. Protect catastrophe and annual totals

52. Read the reinsurance transaction