Your triangle says one thing, but a damage cap passed last year and medical costs are climbing twice as fast as the general economy. Internal history alone will mislead you. Outside information is how you keep the estimate honest.
Standard methods assume the past predicts the future. That assumption fails in three situations. First, the data is thin or volatile: a small line, a new program, or a few large claims swinging the pattern. Second, an environmental change makes old years unrepresentative of future claims. Third, an operational or mix change alters how claims settle. In each case you reach outside the triangle for indices, benchmarks, or legal facts.
KEY: Outside information does one of two jobs. It restates history to today's conditions (trend, on-leveling for reform), or it fills gaps your own data cannot support (benchmark factors, external a priori).
Severities inflate over time. To compare an old claim to a recent one, restate both to a common cost level using a published index.
Common mistakes
- Trending injury losses on general CPI. Bodily-injury severity follows a medical index; using 3% CPI when medical runs 6% understates future cost.
- Extrapolating development across a reform date. Blended factors overstate large-claim development after a damage cap; restate each accident year with a year-specific factor or drop pre-reform years.
- Ignoring the pro rata year. Give the implementation year a full-savings factor and you overstate the reform in that year; only the post-effective-date portion earns the cut.
Bottom line
- Bring in outside data when internal history is thin, volatile, or broken by a structural change (legal, economic, or operational).
- Inflation indices (CPI, medical CPI, wage) restate old claim costs to a common current cost level before you fit a trend.
- Match the index to the loss: medical severity trends on a medical index, wage-loss on a wage index, not on headline CPI.
- Tort reform (damage caps, statute changes) breaks continuity; restate each accident year to a common post-reform environment with year-specific factors, pro rating the implementation year, or drop pre-reform years.
Exam shortcut
When a question mentions medical, wage, or construction costs, reach for the matching index, not general CPI. The exam rewards recognizing that severity type dictates the index. At any legislative change (cap, statute, liability rule), suspect a structural break. The correct move is to restate each accident year with a year-specific reform factor, pro rating the implementation year, or drop pre-reform years, never extrapolate straight across.
The full lesson (about 2,660 words, 18 min read) adds 3 worked examples, all 7 common mistakes, a self-check, free in the app.
Learning objectives
- B20
Browse all free Exam 5 lessons or jump into free Exam 5 practice questions.