Hi, I'm trying to improve my understanding of the calculation of terminal bonuses and links to asset share. I'm aware that often, whole of life terminal bonuses are set based on the bonuses used for equivalent endowment policies. I'd like to understand: 1. Why does this happen? 2. What are the potential issues with this? Thanks
Hi - this a good and interesting question! The issue with whole life assurances is that there is no maturity date. For endowment assurances (let's assume we are talking about conventional with-profits business, so no 'shadow fund' to use), TB rates are set by choosing specimen contracts maturing in the near future, then setting the TB rate to bring the total of {guaranteed benefits + TB} up to the asset share, then smoothing as appropriate - as is described in Section 5.4 of Chapter 19. However, deaths can occur at any time, at any policy duration. For endowment assurances, TB rates payable on death are likely themselves to be based on TB rates for maturities of equivalent durations, with interpolation applied as appropriate. For whole life assurances, we don't have maturing specimen contracts to use to set a TB scale - so using the equivalent rates for endowment assurances is one pragmatic solution. Furthermore, there can be 'problems' with asset shares for whole life assurances. For endowment assurances, the asset share would be expected to increase as the policy progresses towards maturity. However, for whole life assurances asset shares can be inappropriate to use for setting bonuses because, at older ages, the asset shares could be very small or even negative, due to the large 'cost of life cover' deductions. Issues with using the endowment assurance TB rates include: What does the insurer do when the endowment assurances have all matured, and it still has whole life assurances left? For the whole life assurance death payments, the insurer is likely to be paying out more than asset share for some policies and potentially less than asset share for others; if the balance is too far towards the former then this will eat into the estate. [There is actually part of a question on the April 2014 paper which touches on these ideas: see Q1(vii).] Hope that helps.
