Chapter 7 Quiz

Twenty questions to check your grasp of the underwriting decision: fifteen multiple choice and five short answer. Answers are in the collapsed key at the bottom — try the whole set before you open it.

Multiple choice

1. Underwriting is best defined as:

  • A. The process of paying claims fairly and promptly
  • B. The process of evaluating a risk and deciding whether to accept, decline, or modify it, and at what price
  • C. The actuarial construction of rates and reserves from aggregate experience
  • D. The sale of insurance products to prospective customers

2. In the underwriting process, the step that asks "is this risk even ours to quote, and is there a conflict or eligibility problem?" is:

  • A. Triage
  • B. Assessment
  • C. Clearance
  • D. Implementation

3. The chapter argues that most underwriting mistakes are:

  • A. Bad final accept/decline calls made under deadline pressure
  • B. Caused by actuaries supplying inadequate rates
  • C. Failures at an upstream step (thin submission, skipped inspection, missed clearance) rather than the final decision
  • D. The result of overriding predictive models too rarely

4. "Binding authority" specifically refers to the power to:

  • A. Decline any risk for any reason
  • B. Commit the insurer to coverage immediately, putting the company on risk before the policy is issued
  • C. Apply schedule credits and debits to the manual rate
  • D. Set reserves on a reported claim

5. A risk is within your premium and limit authority and in an "acceptable" class, but the building has a hazard your guidelines mark "refer." You should:

  • A. Bind it — you're within your premium and limit authority
  • B. Decline it automatically
  • C. Refer it up the authority chain before binding
  • D. Apply a debit and bind it without referral

6. Authority being "an and, not an or" means:

  • A. You may write a risk if it satisfies any one dimension of your grid
  • B. A risk must satisfy every dimension of your grid; exceeding any single one requires referral
  • C. Two underwriters must jointly approve every account
  • D. You need both binding authority and pricing authority for any decision

7. The four appetite tiers used in the chapter, from most-wanted to never-written, are:

  • A. Preferred, standard, substandard, declined
  • B. Target/preferred, acceptable, restricted/caution, prohibited/declined
  • C. Bind, quote, refer, escalate
  • D. Admitted, surplus, excess, reinsured

8. Underwriting guidelines exist primarily to:

  • A. Replace underwriter judgment entirely
  • B. Translate philosophy and appetite into consistent, defensible selection rules — and to flag what must be referred
  • C. Set the company's investment strategy
  • D. Determine how claims are adjusted

9. "Guidelines encode the past; risks live in the present" is the chapter's way of saying:

  • A. Guidelines are outdated and should be ignored
  • B. No manual can anticipate every atypical account, which is why a referral path and underwriter judgment remain essential
  • C. Underwriters should rewrite the guidelines on every account
  • D. Past loss data is irrelevant to current pricing

10. The three forces an underwriting philosophy must hold in tension are:

  • A. Premium, commission, and surplus
  • B. Frequency, severity, and exposure
  • C. Profit, growth, and social responsibility
  • D. Underwriting, claims, and actuarial

11. In the underwriting–claims–actuarial triangle, the function that sees the aggregate trends and rate adequacy of the whole book (rather than the individual account) is:

  • A. Claims
  • B. Underwriting
  • C. Actuarial
  • D. Distribution

12. The chapter says underwriting "is judgment." The most precise version of that claim is:

  • A. Human judgment always beats data and models
  • B. Every underwriting decision requires deep deliberation
  • C. For complex, novel, and atypical risks, human judgment is irreplaceable; for simple high-volume risks, algorithms write faster and more consistently
  • D. Models should never be overridden

13. A model scores an account 7/10 (decline-leaning), and an experienced underwriter writes it as a 6. The thing the model cannot see, in the chapter's example, is:

  • A. The class code and building age
  • B. The two losses on the loss-summary line
  • C. The catastrophe zone
  • D. The story — that both fires were electrical, predate a new manager, and a roof contract and IR scan are already attached

14. For a model override to count as defensible judgment rather than preference, the chapter insists on:

  • A. Approval from the claims department
  • B. Documented reasoning a skeptical reviewer would accept, and willingness to be measured on whether overrides run better than the model
  • C. A larger premium to offset the extra risk
  • D. A second model that agrees

15. The Harbor Steel account, run through triage in this chapter, is assessed as:

  • A. Target/preferred business to be quoted aggressively
  • B. Prohibited — an automatic decline because the prior carrier non-renewed
  • C. Restricted/caution — writable only with extra controls and pricing, and requiring referral
  • D. Already bound at final terms

Short answer

16. Explain the origin of the word "underwriting" at Lloyd's, and why the chapter says the word's spirit is "exactly right" for the modern job.

17. Give two reasons the underwriting file matters after the decision is made and the underwriter has moved on. For one of them, explain how a thin file can make a good decision look negligent.

18. Why can two equally competent regional carriers rationally place the identical coastal-property risk in opposite appetite tiers (one "restricted," one "target")? What does this reveal about the relationship between appetite and philosophy?

19. Describe "unconscious drift in a soft market" and name one mechanism a disciplined firm uses to force drift to be a decision someone signs rather than a tide no one notices.

20. The override power "cuts both ways." Explain the danger that mirrors the opportunity — how the same judgment that lets a great underwriter rescue a good risk can let a weak one talk themselves into a bad one — and the single discipline that distinguishes the two.


Answer key (try the questions first) **Multiple choice** 1. **B** — Underwriting is the evaluate-and-decide (accept/decline/modify) function, distinct from claims (A), actuarial (C), and sales (D). (§7.1) 2. **C** — Clearance asks whether the risk is yours to quote (conflict/duplicate checks; admitted vs. surplus eligibility). (§7.1) 3. **C** — The decision is usually easy if the upstream steps were done well; the common failure is upstream (thin submission, skipped inspection, missed clearance). (§7.1) 4. **B** — Binding authority is the power to make coverage effective immediately, before the policy issues — the most consequential power an underwriter holds. (§7.3) 5. **C** — A risk that exceeds *any* dimension of your authority (here, a "refer" hazard) must be referred, even if you're within premium and limit. (§7.3, §7.4) 6. **B** — Authority is conjunctive: every dimension must be satisfied; exceeding any one triggers a referral. (§7.3) 7. **B** — Target/preferred, acceptable, restricted/caution, prohibited/declined. (§7.4) 8. **B** — Guidelines turn philosophy and appetite into consistent, defensible selection rules and flag referrals; they do not replace judgment. (§7.4) 9. **B** — The phrase names the limit of any rulebook: atypical accounts need a referral path and underwriter judgment. (§7.4) 10. **C** — Profit, growth, and social responsibility. (§7.2) 11. **C** — Actuarial sees the aggregate (trends, rate adequacy); underwriting sees the individual account; claims sees the consequences. (§7.6) 12. **C** — The precise claim is about *which* risks need judgment (complex/novel/atypical) versus which are better automated (simple/high-volume) — not that judgment always beats data. (§7.7) 13. **D** — The model sees the structured fields (A, B, C); it cannot read the *story* the careful file review reveals (D). (§7.7) 14. **B** — Documented, reviewable reasoning plus accountability for whether overrides actually outperform the model. (§7.7, §7.5) 15. **C** — Restricted/caution, requiring controls, careful pricing, and a referral; not target, not an automatic prohibited decline, and not yet bound. (The Underwriting File) **Short answer (model responses)** 16. At Lloyd's, individuals accepting a share of a marine risk wrote their names *under* the description of the voyage, accepting liability up to a stated amount — to "under-write" was to put your name and money beneath the risk. The spirit fits the modern job because to underwrite is still to personally stand behind a risk: to put the company's promise, and your own professional name, beneath it. (§7.1) 17. Any two of: (a) it lets the decision be *reconstructed* at renewal — why this credit, this deductible, this acceptance of the products exposure; (b) it *defends* the decision under challenge by management, claims, an auditor, a regulator, or a court; (c) it records and tracks the *subjectivities*; (d) it is the *audit trail*. A thin file makes a good decision look negligent because, with no recorded assessment, pricing rationale, or reasoning, there is nothing to show the outcome was a *decision* at all rather than a rubber stamp — "I assessed it carefully" is worth nothing without the documented assessment. (§7.5) 18. Appetite is philosophy made concrete. A carrier whose philosophy favors low-volatility, steady profit will treat catastrophe-exposed coastal property as *restricted* and surround it with controls and referrals; a carrier that has deliberately built a specialty in coastal wind risk — with the cat modeling, reinsurance, and pricing to match — will treat the same class as *target* business it is organized to win. They are not contradicting each other on the *facts* of the risk; they are expressing different deliberate stances toward what risk they want and are equipped to carry. (§7.2, §7.4) 19. Drift is the cumulative, unannounced loosening of price and appetite in a competitive market: a credit here to keep a renewal, a broader appetite there to win an account, marginal risks written "because the market's competitive" — each defensible alone, the sum a book repriced lower and broadened in risk with no one having decided to do either, surfacing as a deteriorating loss ratio two years later. A disciplined firm forces drift to be explicit and signed via any of: an explicit appetite statement, pricing floors, referral thresholds, or underwriting audit. (§7.2) 20. The same judgment that lets a great underwriter see the *fixable* risk the model condemned is the judgment a weak or biased underwriter uses to override a *correct* decline on an account they simply *wanted* — because they like the broker, the account is big, or they're behind on target. The opportunity (rescuing a good risk the model can't read) and the danger (rationalizing a bad risk) look identical in the moment. The discipline that distinguishes them is documented, reviewable reasoning held accountable to results: judgment that cannot be defended — and tested against whether the overrides actually run better than the model — is just preference wearing judgment's clothes. (§7.7)