Your chain-ladder factors assume the past repeats. When claims adjusters tighten case reserves, or the claims department settles faster, or the book shifts toward commercial auto, that assumption breaks and the mechanical answer is wrong. Your job is to spot the shift and correct it.
Development methods assume each accident year runs off the same way. Every development factor is a bet that the emergence pattern is stable. When the operational or economic environment moves, the pattern is not stable, and the factors carry the distortion straight into the ultimate.
KEY: Reported (incurred) methods are sensitive to case reserve adequacy. Paid methods are sensitive to settlement speed. Both are sensitive to inflation and mix. Match the corrective to the moving part.
Before adjusting anything, confirm the environment actually moved. Four ratios do the work.
Read them down each column (across accident years at the same maturity). A rising average case at 12 months signals case strengthening. A rising disposal rate at a fixed maturity signals faster settlement.
Common mistakes
- Trusting a reported chain ladder after case strengthening. The inflated latest diagonal pushes age-to-age factors up; the ultimate comes out too high.
- Trusting a paid chain ladder after settlement speedup. Rising disposal rates make recent years look mature, so the paid method overstates. Do not read the shorter tail as good news.
- Scaling paid in proportion to closed counts. Paid is convex in counts because large claims close late. Multiplying by the disposal-rate ratio understates the restated paid; interpolate the closed-count/paid curve between bracketing cells instead.
Bottom line
- Case reserve strengthening inflates reported (incurred) development factors, so a reported chain ladder overstates ultimate until you correct it; weakening understates.
- Settlement acceleration inflates apparent paid maturity, so a paid chain ladder overstates ultimate; a slowdown understates.
- Before adjusting, consider rearranging the data: swap claim counts for earned exposures, accident year for policy year, report year, or accident quarter, or group by claim size.
- Berquist-Sherman adjusts the data two ways: restate historical case reserves to current adequacy and restate historical paid to one disposal-rate pattern; run both together when both diagnostics move.
Exam shortcut
Read the diagnostic direction to name the bias instantly: rising average case means the reported method is high, rising disposal rates mean the paid method is high. Falling values flip both. Before adjusting, ask whether rearranging the data removes the distortion at the source: earned exposures for claim counts, policy year for accident year, report year under a social or legal shift, accident quarter under rapid exposure growth, or grouping...
The full lesson (about 3,093 words, 21 min read) adds 3 worked examples, all 9 common mistakes, a self-check, free in the app.
Learning objectives
- B11
Browse all free Exam 5 lessons or jump into free Exam 5 practice questions.