Hi, Is there a difference between claim inception date and point of claim date? The notes(2019 version; chp 17; Page 4; point 4) suggest the discount factor is used to discount the annuity value from the point of claim back to policy inception. I think as no payments are being made over the deferred period; so the point of claim is when the claims kick in vis-à-vis claim inception date. Aside from this, do models in real world not consider partial benefits, say if the PH came back to work part time? Is it very tedious to get credible data to estimate this? There is absolutely no reference even when talking about multi state models to a transition to part payments, that could potential be till expiry. Regards, Meher
In addition to the above; Under the LTCI example (2019 version; chp 17; Page 11) the formula only assumes a probability of surviving to time t+1/2. Should we have a probability of eligibility for benefit levels 2 and 3, i.e. the policy holder in questions has survived at the lower level and now has got worse. We need a waiting time in each benefit level no? In fact even the discounting factor wouldn't be the same across these annuities, as the t would change depending on the different times to claim.
Hi Meher This part of the notes is deliberately very brief and vague. There are a wide range of approaches used in practice and so the notes keep it general, rather than specifying a single approach. For example, in the past the course went into great detail about the different types of commutation function that could be used, and in particular looked at two possibilities when the claim date was considered to be either the start of sickness or the end of the deferred period. None of this is examinable any more. Models could include partial benefits and if these are offered they will need to be modelled in some way to price the contract. How sophisticated the modelling can be depends on the data available as you say. Best wishes Mark
Hi Meher Yes, we could make this formula much more complicated to allow for movements between states. There's lots of different ways to do this. For example, the level (1) annuity value could be calculated such that it stops when someone enters level 2 or 3. The claim inception rate for level (2) could then be split into two parts: moving directly to level 2 from healthy, and moving into level 2 from level 1. We then wouldn't need to change the discount factor. As with your previous question, this is not the only way to do it. Best wishes Mark
Adding and referring to Meher's first question- 1.Point 3 is value of annuity payable during the duration of claims- this means these annuity rates includes both discount rates and probability for claim amount till point of claim date. 2.Further in point 4 we are discounting this value from point of claim to policy inception period. (Here I think point of claim above is beginning of claim payments and not claim inception point. While considering cost ,we can assume claim inception can occur anytime.) Now we have cost of claim for entire period of policy at time 0 to determine price with further adjustments. I hope this is the understanding for pricing Income protection. Mark and Meher can add anything more to this concept.
Hi Varsha If we take the point of claim as the point at which claimants start (this is not the only way to do it, but it is a common approach) then yes I agree with your statement on point 4. I'm not very clear about what you say for point 3. The annuity would include the probability of remaining in claim at some future time T and discounting back from T to the point of claim. Best wishes Mark
Yes Mark meant same for point 3 as you stated clearly. Just wanted to specify that discount rates would be included within annuity once claim payment start and we are not multiplying external discount rate. Thanks a lot for such quick and clear replies. Really helpful.
