Non-disclosure, hypertension and a late cancellation notice
On 3 March 2022 Mr Park, then 58, applied for 300 million KRW of whole-life cover through a broker. The application asked whether he had been treated for hypertension in the last five years; he answered no. Pharmacy records later showed amlodipine prescriptions from 2019 onward, and the paramedical examination on 11 March 2022 recorded blood pressure of 148/95 with no comment. Mr Park died of an intracerebral haemorrhage on 2 November 2023. The beneficiary claimed in December 2023. The insurer, which had issued the policy without exclusions, obtained the pharmacy record on 10 January 2024 and posted a cancellation notice on 20 February 2024 citing breach of the duty of disclosure; it also refused the death benefit. The beneficiary has filed suit. The claims manager asks you, as the reviewing actuary, whether the cancellation is effective, whether the death benefit is payable, and what the reserving implication is if the company loses.
1. Is the cancellation effective? Work through the timing rules and the insurer's state of knowledge.
Read the worked answer
No - on these facts the cancellation is out of time.
The right to cancel for breach of the duty of disclosure is subject to lapse periods: one month from the date the insurer knew or could have known of the breach, and three years from the date of contract (Commercial Act Art. 651; confirm the current text and any fraud exception before relying on it in litigation).
- Knowledge: the pharmacy record was obtained on 10 January 2024.
- Notice: posted 20 February 2024 - 41 days later, outside the one-month window. The right lapsed, and a lapsed right is not revived by the beneficiary's claim or by the insurer's later refusal.
- The three-year limit is not the issue here; the contract was under two years old.
There is a second, worse problem. The paramedical examination on 11 March 2022 recorded 148/95. That reading is not by itself a diagnosis, but it is the kind of information that should have triggered a referral, and the insurer will struggle to argue it had no means of knowing. If the underwriting file shows the reading was received and accepted, the argument that knowledge only arose in January 2024 weakens further, and a waiver or estoppel argument becomes available to the beneficiary.
Practical conclusion: the cancellation notice is ineffective, the contract stands, and the insurer should not defend the denial on the cancellation ground.
2. Setting the timing aside, would the non-disclosure have supported a remedy? Discuss materiality and causation.
Read the worked answer
Possibly, but the case is not clean, and the insurer should understand how much it would have won.
Materiality: the test is whether a prudent insurer would have declined the risk, rated it or imposed terms. Treated hypertension with a recorded reading of 148/95 on the day of examination is normally material for whole-life cover at age 58 - a loading or an exclusion would be the expected outcome rather than outright decline. So materiality is probably established, but the remedy would more likely be a proportional reduction or an exclusion than avoidance.
Causation: the death was an intracerebral haemorrhage, which is plausibly connected to hypertension. Korean courts have been willing to limit an insurer's remedy where the undisclosed fact had no connection to the insured event, so the causal link matters for the size of any reduction even where it does not defeat cancellation outright.
The unanswered question - and the one that decides the litigation - is what the underwriter would actually have done. That requires the underwriting manual in force in March 2022 and, ideally, evidence from the underwriter. Without it the insurer is arguing materiality in the abstract.
3. What are the reserving and control implications for the company, whether it wins or loses?
Read the worked answer
Reserving: while the claim is contested, hold the full 300m death benefit as a claim payable, not as IBNR at a discounted probability, because the primary defence is likely to fail on the timing point. Add a separate provision for litigation costs and for interest if the court awards it. Do not run the case on the assumption that a favourable materiality finding rescues the position - it does not, if cancellation has lapsed.
Control implications, which are the more expensive part:
- The one-month clock is an administrative control, not a legal nicety. Ask how many other files had a gap between knowledge and notice, and whether the workflow that logged the pharmacy record also started a timer. If this is systemic, the exposure is the whole cohort, not one policy.
- The broker channel: check the applications submitted by the same broker for the same question and compare with prescription data. A single mis-stated answer is a claims issue; a pattern is a disclosure and mis-selling issue reportable to the FSS.
- The underwriting file itself: a recorded 148/95 accepted without referral is a rule gap. Quantify how many policies in the last three years were issued with an elevated reading and no referral, and consider an in-force review.
- Governance: the appointed actuary should ensure the board sees the aggregate cost of the lapse failures, because remediation of a cohort is far cheaper before the regulator finds it.