IFoA SA2: An IFRS 17 term assurance cohort that has turned onerous

An IFRS 17 term assurance cohort that has turned onerous

Fenwick Life reports under IFRS 17 and groups its term assurance portfolio into annual cohorts of contracts issued in the same year. The 20X1 cohort - 40,000 ten-year level-term policies - was recognised with a contractual service margin of £85m. At the 31 December 20X5 valuation the annual assumption review changes two things: lapse rates fall from 6% to 3.5% a year, and unit maintenance expenses rise 12% after a systems re-platforming. On the revised best-estimate assumptions the cohort is onerous, with a loss component of £22m. Because premiums are level while mortality cost rises steeply with age, the later durations of this product were expected to be loss-making and were previously funded by the assumed lapses. The portfolio is 60% reinsured under a yearly renewable term treaty with an A-rated reinsurer; the treaty is priced annually, and the reinsurer has just quoted a 25% rate increase for 20X6, citing the same lapse experience. The finance director proposes holding the lapse assumption at 6% until the trend is confirmed over another year.

1. Explain how the £22m loss component arises and how it is subsequently accounted for, including what happens if experience later improves.

Read the worked answer

How the loss component arises

The cohort starts with a CSM of £85m. Under the general measurement model, changes in fulfilment cash flows that relate to future service are absorbed by the CSM - but only while the CSM is positive.
The revised assumptions (lapses from 6% to 3.5%, expenses up 12%) reduce the projected profitability of the remaining coverage. The CSM is written down first; the excess is recognised immediately in profit or loss and establishes a loss component on the liability for remaining coverage.
The mechanism must be checked, not assumed. Premiums are level while mortality cost rises with age, so later durations of this product are loss-making by design and were funded by assumed lapses. A fall from 6% to 3.5% keeps policies in force into the durations where they are expected to lose money, and the expense increase compounds it. That is why a lapse reduction is adverse here, which is counter-intuitive and worth stating explicitly in the board paper.
The £22m is therefore highly sensitive to the lapse assumption - which is exactly why the finance director's proposal is a live question.

Subsequent accounting

The loss component is amortised to profit or loss in line with the expected incurred claims and expenses as coverage is provided, so it releases over the remaining term rather than all at once.
It is adjusted for changes in estimates. Favourable changes relating to future service are recognised first as a reversal of the loss component in profit or loss; only once the loss component is exhausted is a new CSM established.
It is tracked and disclosed separately and is not netted against anything else, including the reinsurance result.

The asymmetry is the point

Because the cohort is already in a loss position, further adverse changes in estimates go straight to profit or loss with no CSM to absorb them. Good news is deferred through the loss component; bad news is immediate. This asymmetry belongs in the board's understanding of the earnings profile, not just the accounting notes.

2. The finance director proposes holding the lapse assumption at 6% until the trend is confirmed over another year. Evaluate that proposal and recommend an approach.

Read the worked answer

The proposal as stated is not acceptable - but the underlying question is legitimate

IFRS 17 requires current, unbiased, probability-weighted assumptions reflecting the entity's view of the most likely outcome. Retaining 6% purely to avoid recognising a loss is earnings management, and it would be inconsistent with the assumptions used in pricing and in the capital calculation.
It is also inconsistent with the firm's own commercial reality: the reinsurer has already repriced on a 3.5% lapse view. Retaining 6% would leave the accounts describing an arrangement the firm is not actually buying.

What a defensible response looks like

Challenge the credibility of 3.5% properly. Is the fall broad-based or concentrated in one distribution channel, one intermediary, or an artefact of a retention campaign? Is the sample credible at 40,000 policies? Is the trend visible in the prior two years as well?
If the evidence is genuinely mixed, set a probability-weighted expectation across a documented range rather than ignoring the new data. Record the range, the weights and the reasons. That is unbiased measurement, not deferral.
Check the shape, not just the level. Early-duration lapse behaviour and late-duration lapse behaviour may need separate assumptions, and the anti-selective element at later durations may need its own review.
Separate the expense question. The 12% maintenance expense increase is independent of the lapse debate and should be signed off on its own evidence.
Disclose the sensitivity so the market can see the size of the lapse exposure.
Set a monitoring trigger - monthly lapse experience with a documented threshold for a further assumption review - so the firm is not having this argument again in twelve months.

Recommendation. Recognise the loss on the current best estimate, explain it plainly to the market, and redirect management effort to the two things that actually change the outcome: the retention strategy that caused the lapse fall, and the reinsurance treaty. Adjusting the assumption does not change the cash flows; only those two levers do.

3. The finance director asks whether replacing the yearly renewable term treaty would remove the loss. Advise, explaining how the reinsurance contract held is accounted for and what the realistic options are.

Read the worked answer

The premise is wrong, and it matters

The reinsurance contract held is a separate unit of account, measured separately, with its own CSM or loss-recovery component. Its result is presented separately and does not offset the direct loss component line.
The £22m direct loss has already been recognised on the direct group. Replacing the treaty cannot reverse it; it can only change the future economics.
The 25% rate increase affects the reinsurance contract held: changes in its cash flows relating to future service adjust its CSM, but if the reinsurance contract is in a net cost position or its CSM is exhausted, the effect goes to profit or loss as an additional expense. So the treaty change may worsen reported earnings in the short term even if it improves the economics.

Realistic options

Replace YRT with coinsurance or funds withheld. Aligns the cash flows and timing, may attract better terms because the reinsurer shares more of the underlying experience, but transfers assets and reduces investment flexibility, and may crystallise a cost on day one.
Negotiate a multi-year treaty at a fixed rate to remove the annual repricing exposure. This directly addresses the source of volatility, at the cost of locking in today's pricing and tying up a counterparty for the term.
Reset the retention. Retain more risk and hold the capital instead, if the reinsurer's margin now exceeds the cost of capital.
Recapture and re-tender, but check the recapture terms and any terminal accounting consequences first - recapture can crystallise gains or losses.

How to decide

Model each option on a present-value basis over the remaining term, including new business strain, counterparty default risk, the effect on the internal model and SCR credit, and liquidity. Do not evaluate them on the accounting result alone.

Conclusion for the finance director. No, replacing the treaty will not remove the £22m loss, and any proposal that presents it as a way to do so should be resisted. What a well-structured replacement can do is reduce the volatility of future results and stop the annual repricing risk. Those are good reasons to negotiate - they are just not the reason being given.