Chapter 7 — Key Takeaways

A one-page card. The decision at the heart of insurance, in scannable form.

The core claims

• Underwriting is the decision. It evaluates a risk and answers three questions, in order: Do we want it? At what price? On what terms? — accept, decline, or modify. It is distinct from sales (which wants the business), actuarial (which builds the rate), and claims (which pays the loss).
• Underwriting is a process, not just a moment. Submission → clearance → triage → information gathering → assessment → decision → implementation. The famous decision is usually easy if the upstream steps were done well — and most mistakes are upstream failures, not bad final calls.
• Underwriting is judgment. Data informs and models suggest, but the underwriter decides and must defend the decision. Algorithms write the simple, high-volume risks faster and more consistently; human judgment is irreplaceable for the complex, novel, and atypical ones. The future belongs to those who do both.

The machinery (know these cold)

• Philosophy = the why: profit, growth, and social responsibility held in permanent tension. Profit (the combined ratio) is the floor; growth keeps you relevant; social responsibility is real but cannot survive insolvency. The hardest discipline in insurance is holding price in a soft market.
• Authority = the who: a grid (line · limit · premium · hazard/class · pricing latitude · commitment). It is an and, not an or — exceed any dimension and you must refer. Binding authority (committing the company to coverage immediately) is its sharpest edge; exceeding it can be a firing offense.
• Guidelines = the what/how: the written rulebook encoding appetite and institutional memory. Risk appetite tiers: target → acceptable → restricted/caution → prohibited. Guidelines do the routine work for you — but they "encode the past; risks live in the present," so judgment and a referral path remain essential.
• The file = the record: working tool + professional work product + legal document. Write it so a stranger could defend your decision without you in the room. The decision rationale (what you decided, why, what you required) takes three minutes and matters most on the accounts you're tempted to skip it on.
• The triangle: underwriting (selects/prices) ↔ claims (pays/reserves) ↔ actuarial (analyzes/builds rates). Each sees what the others can't; healthy only when the loop stays an honest conversation, not three departments blaming each other.

The rule of thumb

A risk refers up if it exceeds any single dimension of your authority — limit, class, premium, pricing, or commitment. Being within your limit does not authorize a prohibited class; being in-class does not authorize an over-limit account. Act decisively within your authority; escalate cleanly outside it, with a recommendation attached.

The model-override preview (the book's spine)

A model scores a risk 7/10 (decline-leaning); the underwriter, seeing the story the model can't read, writes it as a 6 with the right terms — and is right to. But the override cuts both ways: the same judgment can rationalize a bad account you simply wanted. The discipline that makes an override judgment rather than preference: documented, reviewable reasoning, held accountable to whether your overrides actually run better than the model. (Paid off in Ch. 31–32.)

Key terms

underwriting · the underwriting process · underwriting authority · underwriting guidelines · binding authority · the underwriting file · risk appetite

The Harbor Steel file this chapter

Run through the front of the process: clean clearance, coastal cat-exposed business. Against appetite it is restricted / caution — not target, not an automatic prohibited decline. The property limit, the named-windstorm exposure, and the loss history each independently push it toward a referral. Disposition: in process; in appetite pending controls and pricing; referral required. What's not settled: whether it is finally a good risk — that needs the information, assessment, math, pricing, and terms of Chapters 8–13.

What you could defend to your manager

"I triaged Harbor Steel as a caution-tier, cat-exposed account, confirmed it's ours to quote, and referred it because the limit, the wind exposure, and the loss history each exceed my grid. I documented the triage conclusion and the open question — can controls and price make this an account we want? — and routed it with my recommendation. I haven't priced or set terms; that's premature until we've gathered the information and assessed the risk."