Close Year

Study guide: Exam 5

Close Year is a data aggregation method that groups claims by the calendar year in which they were finally settled and closed, regardless of when the loss occurred, when it was reported, or when partial payments were made.

\text{Settlement lag} = \text{Close date} - \text{Accident date}

• A close-year cohort is complete the moment it closes — a settled claim has no further development unless it reopens, and reopened claims are a separate component of the unpaid claim estimate (Friedland Ch. 1) — so closed-claim data needs almost no development.
• The trade-off is severe selection bias: within any recent accident year the claims that close first are the small, simple ones. Closed-claim severity therefore understates ultimate severity, and the bias is worst in the long-tail lines where the expensive claims litigate for years.
• Close year is rarely a ratemaking basis. Its real work is operational: measuring the disposal rate (in Friedland, cumulative closed claim counts divided by the selected ultimate claim count for the accident year; other authors define it differently), claim-department throughput, and the average cost of a closed claim.
• Those same closed-claim counts drive the Berquist-Sherman paid adjustment and the disposal-rate variant of the Frequency-Severity Method, where a triangle of closed counts is used to restate paid losses onto a common settlement pattern.
• A shift in closing speed is one of the operational changes an actuary must probe (Friedland Ch. 4): faster closing inflates historical paid development factors, slower closing deflates them — see Claims Processing Changes.

One Claim Under Four Aggregation Bases An auto liability claim arises from an accident on 3/10/2022 under a policy effective 9/1/2021. It is reported 4/1/2022, litigated, and settled 9/14/2024.

Assign the claim under each basis.

Answer

Basis	Year
Policy Year	2021
Accident Year	2022
Report Year	2022
Close Year	2024

The claim sits in a 2022 development triangle for two and a half years, moving along the row as case reserves are revised, and only leaves the inventory in 2024. Close year is the only basis that cannot be assigned until the claim is finished — which is why it is useless for a current-year rate indication and useful for measuring how long claims take.

Closed-Claim Severity Understates Ultimate Severity For accident year 2023, 1{,}000 claims were reported. At 24 months, 700 are closed for total payments of \$3{,}500{,}000; the 300 open claims carry case reserves of \$4{,}200{,}000.

Compare average closed severity to the implied ultimate severity, ignoring further development.

Answer

\begin{align*}
\text{Avg closed claim} &= \frac{\$3{,}500{,}000}{700} \\
&= \$5{,}000 \\[4pt]
\text{Avg open claim} &= \frac{\$4{,}200{,}000}{300} \\
&= \$14{,}000 \\[4pt]
\text{Avg reported claim} &= \frac{\$7{,}700{,}000}{1{,}000} \\
&= \$7{,}700
\end{align*}

Closed claims average \$5{,}000 against \$7{,}700 on a reported basis — and the reported figure is itself understated, since case reserves on litigated claims typically develop upward.

Pricing off closed-claim severity would understate the loss cost by roughly a third. The valid use of the \$5{,}000 figure is as a diagnostic: compare it across accident years at the same disposal rate to detect a real change in claim cost, rather than a change in what has closed so far.