lines of business
A big insurer does not sell one undifferentiated 'insurance'. It sells distinct products: car insurance, home insurance, workers' compensation, general liability, commercial property, and so on. Each of these is a 'line of business' — a grouping of similar policies that cover similar risks, are sold in similar ways, and behave similarly when claims come in. Sorting the company into lines is like sorting a grocery store into aisles: it lets you understand and manage one kind of risk at a time.
Concretely, the main P&C lines include personal auto (your private car), homeowners, workers' compensation (injuries on the job, often legally mandated), commercial property, commercial general liability (CGL — a business's responsibility to outsiders), professional liability and directors-and-officers, and specialty lines like marine or aviation. Lines differ enormously in how they behave: auto claims arrive fast and are mostly small and frequent; liability and workers' comp claims can take years to develop and produce occasional huge payouts. Actuaries almost always analyze each line — and often each coverage within a line — separately, because mixing them would average away the very patterns that matter.
Why it matters: nearly every actuarial calculation in P&C is done by line of business. Pricing, loss reserving, the loss ratio, and capital requirements are all computed line by line, then rolled up. A company can be wildly profitable in homeowners and bleeding money in commercial auto at the same time; only line-level analysis reveals that. Regulators and financial statements also report results by line, so getting the line definitions right is a foundational, not a cosmetic, step.
An insurer's annual report shows personal auto with a 95% loss ratio, homeowners at 70%, and commercial liability at 130%. Same company, three lines, three very different stories — the liability line is losing money and needs a price increase or tighter underwriting.
Each line behaves differently, so actuaries analyze and report results line by line.
Lines are not standardized worldwide — what one country calls 'casualty' another splits into several classes. The right granularity is a judgment call: too coarse hides patterns, too fine leaves too little data per group to be credible.