Underwriting is the process an insurer uses to evaluate the risk of a potential policyholder, decide whether to accept the risk, and set the premium and terms — the discipline that determines whether the business collects enough premium to cover the losses it will eventually pay.
Why It Matters
Underwriting quality is the difference between an insurer that survives a bad accident year and one that doesn't — pricing risk too low to win market share is a well-worn path to insolvency, and it's why regulators scrutinize reserve adequacy so closely. It's also the function most actively being reshaped by AI-assisted risk scoring and alternative data, since faster, more granular underwriting is a genuine competitive differentiator.
How It Works in Practice
- 1Gather applicant data: financial, health, property, or driving history depending on the line of business
- 2Assess risk against underwriting guidelines and actuarial models, classifying the applicant into a risk tier
- 3Price the policy (premium) to reflect that risk tier, factoring in expected losses, expenses, and target profit margin
- 4Decide: accept at standard terms, accept with modified terms (exclusions, higher deductible, higher premium), or decline
Common Pitfalls
Underwriting to win volume in a soft (competitive) market without adequately pricing for risk is a classic cause of later reserve deficiencies
Adverse selection — the tendency for the highest-risk applicants to seek out the most generous coverage — punishes underwriting models that don't segment risk finely enough
AI-assisted underwriting models trained on historical data can encode and perpetuate biases in ways that create real regulatory and fair-lending/fair-pricing exposure