Coverage Units

Study guide: Exam 6C

Coverage Units are the IFRS 17 measure of the quantity of service a group of contracts provides in a period, used to allocate the Contractual Service Margin to revenue. Each period's CSM release is the CSM available divided across the coverage units provided this period and expected in future periods.

\text{CSM released} = \text{CSM} \times \frac{\text{Coverage units this period}}{\text{Total coverage units remaining}}

• What a coverage unit is: the quantity of coverage provided, determined by considering the amount of benefits the contracts could provide and their expected duration. IFRS 17 does not prescribe a formula — the insurer selects a driver that reflects the service and applies it consistently.
• Common drivers: the number of contracts in force (simplest, appropriate where each provides similar coverage), sum insured or policy limit (where amounts differ materially), expected claims, or time. The choice is a significant judgement and must be disclosed.
• Expected duration is included, so coverage units must reflect expected lapses — units are counted for the coverage actually expected to be provided, not for the full contractual term ignoring cancellation.
• Revised each period. The pattern of expected future coverage units is updated, which changes the release rate prospectively. A group whose expected duration shortens releases CSM faster.
• Not applicable under the PAA for the LRC, since there is no CSM to release. Coverage units matter to a Canadian P&C insurer mainly on GMM-measured multi-year contracts and on some reinsurance held.
• The choice can change reported profit substantially: a limit-weighted driver front-loads profit on a group with large early-expiring policies, while a contract-count driver spreads it evenly.

Choosing the Coverage Unit Driver A three-year group contains two sub-blocks: 1{,}000 contracts with a \$1 million limit each, all running the full three years; and 4{,}000 contracts with a \$100{,}000 limit each, expected to lapse evenly so that only 50\% remain in year 2 and 25\% in year 3. The CSM at inception is \$15 million.

Compute the year-1 CSM release under a contract-count driver and under a limit-weighted driver.

Answer Contract count. Contracts in force: year 1, 1{,}000 + 4{,}000 = 5{,}000; year 2, 1{,}000 + 2{,}000 = 3{,}000; year 3, 1{,}000 + 1{,}000 = 2{,}000. Total 10{,}000.

\begin{align*}
\text{Year 1 release} &= \$15\text{M} \times \frac{5{,}000}{10{,}000} \\
&= \$7.5\text{M}
\end{align*}

Limit-weighted. Limits in force, in millions: year 1, 1{,}000 + 400 = 1{,}400; year 2, 1{,}000 + 200 = 1{,}200; year 3, 1{,}000 + 100 = 1{,}100. Total 3{,}700.

\begin{align*}
\text{Year 1 release} &= \$15\text{M} \times \frac{1{,}400}{3{,}700} \\
&= \$5.68\text{M}
\end{align*}

A \$1.8 million difference in year-1 profit from a single judgement, with no difference in cash, claims or coverage.

Which is right? The limit-weighted driver, on these facts. The large-limit contracts provide far more coverage per contract, and they persist while the small ones lapse — so counting contracts treats a \$100{,}000 policy as equal service to a \$1 million one, which does not reflect "the amount of benefits the contracts could provide." The contract-count driver would be defensible only if the limits were similar.

This is why the coverage-unit driver must be disclosed: it is a lever on reported profit, and a reader cannot interpret the CSM release without knowing which one was pulled.