Underwriting analytics

Underwriting analytics is the data analysis used to decide whether to accept a specific risk — an insurance policy application or a loan request — and at what price. It draws on the applicant's own data, third-party sources such as credit bureaus or property records, and, increasingly, alternative data sources to assess risk more granularly than a small set of fixed rating factors could.

In insurance, underwriting analytics scores an individual application's expected loss and sets premium accordingly, drawing on the same probability estimates that actuarial modeling develops for the book of business as a whole; in lending, the equivalent process is closely tied to credit scoring and default-probability estimation. Both differ from claims or loan-servicing analytics, which operate after the underwriting decision is made, and from portfolio-level risk adjustment, which normalizes comparisons across an already-accepted population rather than deciding whether to accept it.

Underwriting analytics matters because pricing risk accurately — neither too high, which loses good customers to competitors, nor too low, which invites adverse selection and drags down the loss ratio and combined ratio — is central to profitability in both industries. A common pitfall is over-relying on a single predictive variable that correlates with risk in historical data but breaks down when market conditions shift, or that inadvertently proxies for a protected characteristic and creates fair-lending or fair-pricing exposure.