K-ICS and IFRS 17 after a rate move: which number drives the dividend
Saehan Life reports K-ICS own funds of 4.20 trillion KRW against an SCR of 2.50 trillion, a ratio of 168%, and IFRS 17 CSM of 1.10 trillion, of which 0.35 trillion sits in a whole-life block carrying a 3.0% minimum crediting rate. Liability duration is 14 years; 60% of invested assets are government bonds of duration 6, 30% A-rated corporate bonds of duration 5 and 10% is property and equity. In the third quarter the 30-year KTB yield rose 45 bp and A-rated spreads widened 70 bp. The company uses the OCI option for insurance finance income and expense. To defend the ratio ahead of the year-end FSS review, management proposes three actions: cede 30% of a profitable protection block on coinsurance with an 8% ceding commission; re-cut coverage units to release 150 billion KRW of CSM; and apply the FVOCI bond revaluation loss against CSM so that reported equity is unaffected. The board asks the appointed actuary to separate what is legitimate from what is not, and to say which number should drive the dividend decision.
1. Assess each of the three proposals: is it legitimate, and what is its actual accounting and capital effect?
Read the worked answer
(a) Cede 30% of a profitable protection block. Legitimate in principle, but it is a real economic transaction, not a capital trick. On coinsurance the ceded share of the CSM is derecognised, so CSM falls by roughly 30% of that block's CSM, and own funds fall by the present value of the profit given away. SCR falls by the mortality, morbidity and lapse risk on the ceded share plus some diversification benefit. The ratio improves only if the SCR relief exceeds the own-funds reduction - which is plausible for a protection block with high lapse and mortality risk, but must be demonstrated, not assumed. The 8% ceding commission must be at market; if the treaty contains a side letter or the reinsurer is a related party that returns the risk, the risk transfer fails and the accounting follows the substance.
(b) Re-cut coverage units to release 150bn of CSM. Not legitimate if done retrospectively. Coverage units must reflect the quantity of benefits provided in each period; a change is permitted only for a genuine prospective change in the service delivered, from the date of change, applied to the CSM remaining. Changing the pattern to accelerate release is an accounting change dressed as an assumption change, and the auditor and the FSS will treat it as such.
(c) Apply the FVOCI bond loss against CSM. Cannot be done. CSM is an unearned profit on the liability side, not a revaluation reserve. Asset-side OCI cannot be netted against it; doing so would defer a loss that has already occurred and misstate both equity and the CSM roll-forward.
2. Explain how the K-ICS ratio can improve while IFRS 17 equity is under pressure, and show the sensitivity arithmetic.
Read the worked answer
The two frameworks measure the liability differently. K-ICS discounts all best-estimate cash flows at current rates and adds a risk margin at a high confidence level; IFRS 17 uses a risk adjustment at a lower confidence level and, under the OCI option, splits insurance finance into a locked-in amount in profit or loss and the discount-rate difference in OCI. CSM is not a component of K-ICS own funds at all, so a large CSM is no comfort to the supervisor.
Sensitivity arithmetic on the quarter's moves:
- Liabilities: duration 14 x 45 bp = about -6.3%.
- Government bonds (60%, duration 6): -6 x 0.45% = -2.7%.
- Corporate bonds (30%, duration 5, spread +70 bp): -5 x 1.15% = -5.75%.
- Weighted asset change: 0.6 x (-2.7%) + 0.3 x (-5.75%) = -3.35%.
If assets and liabilities are roughly matched in size, the liability falls more than the assets, so net own funds rise by about 3% of the balance and the ratio improves. Under IFRS 17, the same rate rise pushes the FVOCI asset losses through OCI, while the liability gain also lands largely in OCI under the OCI option - so IFRS equity can move the other way, and the CSM is barely touched by a pure discount-rate move.
Conclusion: the ratio improving is not evidence that economic solvency improved. It is a duration gap being measured twice.
3. The board asks which number should govern the dividend decision. Give your recommendation and the checks you would require first.
Read the worked answer
Neither the K-ICS ratio alone nor IFRS 17 equity alone. Dividends require distributable profits measured under IFRS (retained earnings and, for a Korean company, the statutory and capital-reserve constraints), and they require the company to remain above its K-ICS and any supervisory target after the payment. The binding constraint is the lower of the two, tested after a stress.
Checks before any recommendation:
- Recompute the K-ICS ratio on a pro-forma basis after the dividend, not before it, and after the SCR impact of the coinsurance treaty is recognised.
- Test the ratio under a defined stress - a 100 bp fall in rates with the same duration gap - because the current improvement came from a rate rise and can reverse.
- Confirm whether the FVOCI losses have already reduced retained earnings: under IFRS 17 equity includes OCI, so unrealised bond losses do constrain distributable profit even though they do not change CSM.
- Establish the CSM release for the next four quarters from the existing coverage units, and confirm that this is the only CSM release supporting the dividend; a CSM-based dividend funded by an accelerated release is not sustainable.
- Document that the coinsurance is at arm's length, with the gain or loss on reinsurance recognised in profit or loss as required.
If the pro-forma post-dividend ratio under stress stays above target and IFRS retained earnings support the payment, proceed - but at a fraction of what the pre-stress 168% appears to allow.