Chapter 34 — Key Takeaways
A one-page field card for InsurTech and the digital transformation of insurance.
The core claims
- InsurTech is not one thing. Sort every player by what part of the value chain they attack: full-stack carriers (own the risk), digital MGAs (underwrite on a carrier's paper for a fee), enablers (sell software to incumbents), and distributors/embedded platforms (the front door). The most important question about any of them: whose balance sheet carries the loss?
- The durable winners are capital-light. Digital MGAs and embedded platforms added technology to distribution and left the carrying of insurance risk to balance sheets built for it. They aged better than the full-stack carriers that took on the whole chain — including the loss ratio.
- The full-stack bet ran aground on the loss ratio. Many listed InsurTech carriers optimized the expense ratio and assumed it would deliver the combined ratio. But combined = loss + expense, and the loss ratio is the larger, harder term. Only underwriting moves it.
- Embedded insurance relocates underwriting upstream. It becomes class underwriting (Chapter 20) of a whole population; the work moves into program design, and an error there is made automatically thousands of times.
- Parametric and usage-based products are real, with real limits. Parametric pays on a trigger, not a loss — fast and cheap to settle, but carrying basis risk and requiring insurable interest (Chapter 4) to be designed in. Usage-based cover attacks adverse selection by replacing a proxy with measurement — but it is a data relationship with data-quality, privacy, and fairness questions (Chapter 35).
- The underwriter's job distilled, it did not vanish. Routine risks automate (and good riddance); the complex-judgment work and the new work of building, supervising, and governing the automated systems became more valuable. The winner does both craft and technology.
The rule of thumb
Growth is easy; underwriting is hard. A low expense ratio cannot rescue a high loss ratio. You can buy market share in insurance with a click — drop the price, loosen the rules, say yes more — and adverse selection guarantees the eager buyers skew bad. Only selection and adequate pricing move the loss ratio, and the combined ratio always tells the truth eventually.
The three questions to ask of any "disruptor"
- Whose balance sheet carries the loss? (Insurance risk vs. distribution risk are different businesses.)
- What is the loss ratio, not the growth rate? (Growth is purchasable; profit is not.)
- What did they actually automate — intake or judgment? (Automating paperwork is real; automating the decision on a complex risk is what keeps blowing up.)
Key terms
InsurTech · embedded insurance · digital MGA · peer-to-peer insurance · API distribution (plus, used from their owners: MGA, Ch. 3; STP/class underwriting, Ch. 20; parametric, Ch. 26; usage-based insurance, Ch. 14; combined/loss/expense ratio, Ch. 3; insurable interest, Ch. 4.)
What you could defend to your manager
"A digital MGA's quote-in-seconds platform could speed the intake on Harbor Steel — but a sound referral rule should refuse to auto-bind it and send it to a human, because the two fires, the aging roof and sprinklers, the named-windstorm exposure, the pending products claim, and the multi-line structure are exactly the complexity that requires judgment. The machine is for the small clean machine shop down the road, not for this. A parametric wind supplement is worth raising with the broker as a fast-liquidity option over the 5% named-windstorm deductible — but only as a supplement, with its basis risk and insurable-interest limits disclosed. The account stays referred and human-underwritten."