Chapter 2 — Key Takeaways

A one-page card. The history is not antiquarianism: every form, market, and formula on your desk is the residue of a problem someone solved — usually after a disaster.

The core claims

• Risk transfer is three thousand years old. Insurance did not begin with insurance; it began with merchants trying not to be ruined by a single venture, and with two ancient inventions that are its direct ancestors: the bottomry loan (transfer the loss for a premium-like charge) and general average (share a deliberate sacrifice among all who benefit — risk pooling as a legal rule, still alive in marine policies).
• The Great Fire of London (1666) birthed fire insurance and three permanent levers. Out of the rebuilding came the first true fire offices, which invented (1) selection/pricing by physical characteristics (timber vs. brick → construction class and rating factors), (2) inspection before acceptance, and (3) prevention as the insurer's business (the fire mark and private brigades → loss control). You still pull all three levers on every account.
• A coffeehouse became a market — and named the profession. At Lloyd's of London (origins), individuals wrote their names under a marine risk to accept a share — hence underwriter — creating the subscription market (ancestor of co-insurance, layering, reinsurance) and the broker-underwriter structure you work inside every day. Lloyd's is a marketplace of syndicates, not an insurance company.
• The mortality table tamed life insurance and created the actuary. Halley's life table showed death could be counted in the aggregate; Equitable Life (1762) priced by age from mortality data and held adequate reserves, converting life insurance from a wager into a discipline and triggering the rise of actuarial science. The actuary prices the class; the underwriter prices the case.
• Insurance reforms by catastrophe. The tontine/deferred-dividend scandals drove major life-insurance reform and stronger state regulation; insurer failures drove state-based regulation (secured by McCarran-Ferguson, Ch.4); Hurricane Andrew, September 11, and Katrina reshaped catastrophe pricing, deductibles, and reinsurance. Every form is a memorial to a loss someone absorbed first.
• From ledger to algorithm is one long arc, not a break. Ledger → rating bureau/manual (ISO, NCCI) → statistical models → computers → multivariate models (GLMs) → machine learning on satellite/IoT/telematics data. Every tool does the same job (estimate the risk, suggest a price, ever more finely) and shares the same blind spot (it prices the past and the class; it cannot see the particular, novel, or changing thing about the individual risk). The tool advises; the underwriter decides.

The rule of thumb

Insurability is a property of the risk plus the machinery available to manage it. Fire was "uninsurable" on land in 1600 and routine by 1700 — the peril never changed; the pool, the data, and the tools did. Apply this to every "uninsurable" modern risk (cyber, climate, pandemic): usually the concept exists; the pool and the data do not yet.

The continuity to remember

Key terms

bottomry · general average · Lloyd's of London (origins) · tontine · the rise of actuarial science

What you could defend to your manager

"The percentage wind deductible, the catastrophe load, and the cession to reinsurance on Harbor Steel's coastal exposure are not arbitrary — they're the distilled memory of storms like Andrew and Katrina that ruined other carriers so ours could learn to price the risk. And the model's decline-leaning score is the latest in a three-thousand-year line of tools that advise; reading the particulars it can't see, and overriding it when the file warrants, is the underwriter's job and always has been."