Chapter 33 Exercises

Work these with the chapter's two habits of mind: of every discrepancy ask is it material, and is there intent?, and of every red flag ask what is the innocent explanation, and is this one flag or a cluster? Items marked with a dagger (†) have worked solutions in Appendix: Answers to Selected Exercises; the rest are for discussion or self-test. Section references like (§33.4) point you back to the relevant part of the chapter. Throughout, remember the bright line of this entire chapter: a red flag is a prompt to investigate, never a finding of fraud — write your answers in that register.

A. Recall and definitions

  1. † Define insurance fraud and name the three elements the chapter pulls out of the definition (deliberate, material, financial gain). Why must all three be present? (§33.1)
  2. Distinguish soft fraud from hard fraud in one sentence each, and say which is more common and which is more costly per instance. (§33.2)
  3. † Define material misrepresentation and list its three elements (false statement of fact, materiality, reliance). What is the classic test for materiality? (§33.3)
  4. Define concealment and explain the one element it requires that misrepresentation does not, and why that makes concealment harder to prove. (§33.3)
  5. Define rescission in one sentence, and explain how it differs from both non-renewal and a coverage denial. (§33.4)
  6. Define a red-flag indicator, and complete this rule in your own words: "one flag is a ; a cluster across families is a ." (§33.5)
  7. † Define the special investigation unit (SIU) and state the division of labor between the underwriter and the SIU in the fraud-response chain. (§33.6)
  8. Name the three families of fraud-detection analytics in §33.7 and say, in a phrase, what each one finds. (§33.7)

B. Soft, hard, and the gray zone

  1. For each of the following, place it on the fraud spectrum (honest error / gray zone / soft fraud / hard fraud / organized) and justify the placement in one line: (a) an applicant forgets a \$700 claim from five years ago; (b) an applicant rounds payroll from \$11.3M down to "about \$11M"; (c) a claimant adds a never-owned laptop to a real burglary list; (d) a ring stages collisions across dozens of policies. (§33.2)
  2. † Explain why soft fraud is harder to detect than hard fraud, and why much of its cost ends up simply absorbed into a class's loss ratios — and what that does to the honest insureds in the class. (§33.2, §33.1)
  3. The chapter warns against treating fraud as a binary "yes/no light" rather than a "dimmer switch." Give a concrete example of the error an underwriter makes who sees fraud as binary, and the cost of that error. (§33.2)

C. Misrepresentation and materiality

  1. † For each omission, judge whether it is plausibly material and what your next action is: (a) the applicant misstated the building's year of construction by one year; (b) the applicant did not disclose a \$300,000 fire two years ago; (c) the applicant listed the operation as "light assembly" when it includes daily welding. (§33.3)
  2. A misrepresentation must be a false statement of fact. Explain why "I expect to hire ten more employees next year," which turns out false, is generally not a misrepresentation, and construct one statement on the same application that would be. (§33.3)
  3. † The chapter says "utmost good faith runs both ways" and that "good fraud defense begins with a good application." Explain how a vague or ambiguous application question can defeat the insurer's later claim of misrepresentation, and what this implies for how you (or your product team) should draft questions. (§33.3, Ch.4)
  4. Why does establishing that an application answer was false tell you nothing, by itself, about whether it was fraud? Name the element that the falsehood alone does not establish, and the only ways to establish it. (§33.3)

D. Rescission and its limits

  1. † List four distinct limits on the insurer's power to rescind (from incontestability, innocent misrepresentation, waiver/estoppel, varying state standards, the burden of proof). For each, state in one line how it could defeat a rescission. (§33.4)
  2. Define post-claim underwriting and explain precisely why courts and regulators disfavor it. What is the disciplined alternative, and which earlier chapter supplies the tools for it? (§33.4, Ch.8)
  3. † Underwrite this scenario. A homeowner suffers a kitchen fire. While adjusting the claim, the insurer discovers the application understated the home's square footage by 8% — a fact unrelated to the fire. The claims manager wants to rescind to avoid the payment. As the referral underwriter, write your recommendation: should the carrier rescind? Address materiality, the post-claim-underwriting risk, and the litigation exposure. (§33.3, §33.4)
  4. Explain why a rescission that cannot be proven is "worse than no rescission." What new exposure does a failed rescission create? (§33.4)

E. Find the red flag

  1. † Find the red flag(s). Read this submission and list every red flag, grouping each by its family (§33.5), then give the most likely innocent explanation for each, and finally state whether the overall picture is a question or a referral — and why:

    text SUBMISSION — "Crescent Auto Detailing & Customs" [constructed teaching example] Requested: commercial property + GL, bind date TOMORROW. Building values requested: well above the modest leased-shop operation described. Loss history: "none" on the application; prior carrier named, loss runs "to follow." Entity: formed 4 months ago; two prior business addresses in 18 months. Financials: owner mentions the business "has been slow" and is "behind on the lease." Inspection: owner asks to "hold off" on the physical inspection until after binding.

  2. Take exercise 20 and change exactly one fact so that the picture flips from "referral" to "ordinary account, one routine question." Which fact did you change, and why does removing it collapse the cluster? (§33.5)

  3. † Write the neutral, documented question you would send the broker about the "no prior losses" answer in exercise 20, given that your loss-run search later shows a \$25,000 claim. Then explain why a second false explanation, if it came back, would be more damning than the first omission. (§33.5)
  4. Explain why the innocent-explanation column of the red-flag table is a genuine part of the detection method and not mere courtesy. What does an underwriter who skips it get wrong, statistically? (§33.5)

F. The SIU and the referral

  1. † Write the referral. Using the Harbor Steel disclosure gap (the 2023 fire's cause was shaded) as your facts, draft the five components of a good fraud-or-disclosure referral note from §33.6 — or explain, with reasons, why this particular gap is a clarification rather than an SIU referral. (Either answer can earn full credit if reasoned correctly.) (§33.6, The Underwriting File)
  2. The chapter says "most referrals come back cleared" and treats that as a feature, not a failure of the system. Explain why, and what it implies for the underwriter who is reluctant to refer "because I don't want to accuse anyone." (§33.6)
  3. Explain the two opposite failure modes in referring to the SIU — under-referring and over-referring — and the harm each does to the unit's ability to catch real fraud. (§33.6)
  4. † Compliance. Name three distinct legal duties or constraints that bind fraud investigation (from: unfair-claims-practices timelines, defamation exposure, good-faith reporting immunity, FCRA/privacy limits). For each, state the disciplined practice that keeps the underwriter on the right side of it. (§33.6, Ch.4, Ch.8)

G. Data and analytics

  1. † Match each technique to what it is best at and to its characteristic false positive: anomaly detection, link analysis, predictive fraud models. (§33.7)
  2. Explain, using the Chapter 32 framework, why a high fraud score is "a strong prompt to investigate" but cannot, by itself, deny a claim or rescind a policy. What is the fraud-side analogue of the documented override? (§33.7, Ch.32)
  3. † Ethics dilemma. Your carrier deploys a predictive fraud model trained on the last ten years of SIU investigations. An analyst notes that the model flags claimants from one set of ZIP codes at sharply higher rates. A colleague says, "the model just found where the fraud is." Write a one-paragraph response: what is the danger the colleague is missing, which chapter owns the full treatment, and what should the carrier do before trusting the model to triage? (§33.7, Ch.32, Ch.35 preview)
  4. Link analysis is described as "most lethal to fraud rings." Explain why a ring is vulnerable to link analysis specifically, even when each individual policy looks innocent — and give one way a link can be an honest coincidence. (§33.7)

H. Memo and integration

  1. † Write the memo. You are the underwriter on a mid-size commercial submission. A single material discrepancy surfaced (an undisclosed but real prior loss), the applicant gave a plausible written explanation, and there is no cluster of other flags. Write a short decision memo (six to ten sentences) to your file documenting: the flag, the action you took, the explanation received, your materiality and intent assessment, the disposition, and why you did not refer to SIU or pursue rescission. (§33.3, §33.5, §33.6)
  2. Price this risk (qualitatively). Two otherwise-identical commercial classes have the same reported loss ratios, but you have reason to believe Class A's losses are inflated by widespread soft fraud while Class B's are clean. Explain why charging both classes the same "adequate" rate is a problem, who is overpaying, and what terms (not just price) you could use to push back on the soft fraud. (§33.2, §33.1, Ch.11, Ch.12)
  3. † Underwrite this submission. A new commercial-property account arrives over-insured relative to its operation, with the prior carrier non-renewing and the loss runs not yet provided, wanting to bind in 48 hours. Lay out your decision process: what you require before binding, which red-flag families are present, what would move this from "writable with verification" to "decline or refer," and how you avoid both rubber-stamping and false accusation. (§33.5, §33.6, Ch.8, Ch.13)
  4. The Underwriting File — extension. Suppose that, instead of merely shading the 2023 fire's cause, the Harbor Steel application had concealed the 2023 fire's existence entirely, and your loss-run order revealed it. Rewrite the Underwriting-File disposition for this counterfactual: where does it now sit on the fraud spectrum (§33.2), is it a referral, is rescission now on the table (§33.4), and what do you do before binding (if you bind at all)? Be explicit about what changed from the real file and why. (The Underwriting File, §33.2, §33.4, §33.6)
  5. Integration across the book. The chapter calls fraud "adverse selection in its most deliberate form." Trace one thread: pick a single underwriting tool from an earlier chapter (the application, the loss run, the inspection, the deductible, or the model) and explain how the same tool defends against both ordinary adverse selection (Ch.1) and deliberate fraud — and where the tool's defense against fraud is weaker than its defense against honest bad risk. (§33.1, Ch.1, Ch.8, Ch.12)