Question ID: 2558
Regulation Reference: (EU) No 2015/35 - supplementing Dir 2009/138/EC - taking up & pursuit of the business of Insurance and Reinsurance (SII)
Topic: Solvency Capital Requirement (SCR)
Article: 138(2)
Status: Final
Date of submission: 12 Jan 2023
Question
For longevity risk it is stipulated that "The decrease in mortality rates referred to in paragraph 1 shall only apply to those insurance policies for which a decrease in mortality rates leads to an increase in technical provisions without the risk margin".
In an answer to a previous question (462) it has been clarified that these technical provisions refer to the BEL gross of reinsurance.
Another question (1523) queries whether the decrease in BEL should be measured where FDB is fixed (BSCR) or where FDB is allowed to change (nBSCR). The answer given here is that both of these options should lead to the same subset of policies having a decrease in BEL. In this answer it is also stated that "8. In case information should become available that there are cases in which the situation described in paragraph 6 reflects the actual risk sharing of the insurance contract, the issue will be revisited and this answer possibly revised."
We have a portfolio of policies with a shared smoothing mechanism between policies. For this portfolio there are policies where the BSCR can increase while the nBSCR decreases, or vice-versa. This happens as some policies might receive a bit more of the shared loss-absorbing buffer in a stress scenario than they did in the base scenario, even if their BEL without FDB increases. This could then realistically lead to BSCR>0 and nBSCR0 could then happen for other policies. For the portfolio as a whole the situation would not arise as the smoothing only happens between the policies of the portfolio.
This means that applying the condition to individual policies will lead to a misstatement of the stress for the portfolio as a whole.
In this situation, how should the requirement be handled? Should it be applied to the closed portfolio as a whole, in order to capture and not overstate the smoothing mechanism?
EIOPA answer
As Article 138 of Delegated Regulation 2015/35 (DR) on the longevity risk sub-module is phrased in an analogous manner to Article 137 DR on the mortality risk sub-module, the considerations of Q&A 462 can indeed be transferred to longevity risk. That means the change in mortality rates shall in both cases be applied only to those policies for which this change leads to an increase of technical provisions without the risk margin and gross of reinsurance.
Furthermore, in general the same policies should be selected when assessing the impact of the stress excluding and including changes to the future discretionary benefits (FDB).
The following answer is based on the assumption that the smoothing mechanism relates to the determination of FDB.
As per Article 206(3) DR, the calculation of FDB should reflect all legal, regulatory or contractual restrictions. This means that the full impact of the smoothing mechanism on all policies should be taken into account in that calculation.
Therefore the sequencing that should be followed for longevity risk is:
- Step 1: apply the longevity shock to all policies and calculate the best estimate gross of reinsurance, excluding changes to the FDB, and thus also excluding the smoothing mechanism
- Step 2: select those policies where the longevity shock has increased the best estimate
- Step 3: perform a scenario where the longevity shock is applied only those policies
- Step 4: take into account the impact of the scenario on FDB including the smoothing mechanism in relation to all policies
- Step 5: calculate BSCR from Step 3 and nBSCR from Step 4.
Where the result of Step 1 and Step 3 calculations need to be estimatedbased on fixed FDB because changes of FDB are an integral part of the valuation model, this estimation should be performed in a prudent and proportionate manner.
The same approach should be applied to mortality risk.
Q&A 1523 will be revised accordingly.