Chapter 18

[rlsrachaellouisesmith]

On page 7 of the notes it states that smoothing profits/losses can arise because of smoothing.

We expect smoothing profits/losses to be neutral over time when applied to all benefits paid.

Is it also true that we expect the cost of smoothing to be neutral over time for smoothing of surrender values?
I don't think it is true, because they are only part of the aggregate of all benefits paid. So we may see profits/losses on the surrender values and losses/profits on the maturity/death claim values.

Or when we are talking about smoothing of surrender values, are we talking about smoothing over the term of the policy, so early surrenders lead to a higher surrender value than can be afforded, and later surrenders lead to a lower surrender value than can be afforded, due to the need to smooth the SV to the maturity value.

Additionally, there are two approaches to addition of surrender p/l cashflow and addition via inv return. In the addition via inv return it says
this way, profits can be allocated over a wider range of policies and product types.
This appears to be suggesting that the cashflow addition method is not suitable for this - is this the case or am I misinterpreting?

Thank you

 

[Lindsay Smitherman]

On page 7 of the notes it states that smoothing profits/losses can arise because of smoothing.

We expect smoothing profits/losses to be neutral over time when applied to all benefits paid.

Is it also true that we expect the cost of smoothing to be neutral over time for smoothing of surrender values?
I don't think it is true, because they are only part of the aggregate of all benefits paid. So we may see profits/losses on the surrender values and losses/profits on the maturity/death claim values.

I'm not sure I agree with this. When setting surrender values, the company can do so on the basis that it expects the smoothing to be cost-neutral over a long enough period of time. However, bear in mind that there might well be less smoothing of payouts for surrenders, in order to avoid anti-selection: so surrender values may be closer to unsmoothed asset share than to smoothed asset share - as per the use of MVRs on UWP business.

 

[Lindsay Smitherman]

Or when we are talking about smoothing of surrender values, are we talking about smoothing over the term of the policy, so early surrenders lead to a higher surrender value than can be afforded, and later surrenders lead to a lower surrender value than can be afforded, due to the need to smooth the SV to the maturity value.

In this context, smoothing is normally referring to smoothing out investment return fluctuations: paying higher than (unsmoothed) asset share under poor investment conditions, lower than asset share under strong investment conditions. [The term 'smoothing' could also refer to smoothing across policies, given that the same bonus rates apply to a cohort of policies, rather than each policy having its own rate, hence there being some inevitable cross-subsidies.]

 

[Lindsay Smitherman]

Additionally, there are two approaches to addition of surrender p/l cashflow and addition via inv return. In the addition via inv return it says
this way, profits can be allocated over a wider range of policies and product types.
This appears to be suggesting that the cashflow addition method is not suitable for this - is this the case or am I misinterpreting?

Yes I think you might be misinterpreting it. The cashflow addition approach allows more tailored addition of any surrender profits to policies, reflecting the experience within that cohort. (Although it could alternatively be done across a wider range of policies - as we explain in the second paragraph under the heading 'Cashflow addition to asset share' in Section 3.3 of Chapter 18.) The investment return method could be applied by just giving the same additional return to a wide range of policies / products, hence everyone gets the same amount - more broadbrush, less tailored, less reflective of the experience within particular cohorts. However, the latter could alternatively be done by cohort, as we explain in the second paragraph under the heading 'Addition via investment return'.

 

[rlsrachaellouisesmith]

I'm not sure I agree with this. When setting surrender values, the company can do so on the basis that it expects the smoothing to be cost-neutral over a long enough period of time. However, bear in mind that there might well be less smoothing of payouts for surrenders, in order to avoid anti-selection: so surrender values may be closer to unsmoothed asset share than to smoothed asset share - as per the use of MVRs on UWP business.

Ok, thank you.

So, for maturity and death benefits might be greater levels of smoothing so payouts are closer to smoothed earned asset share.

Whereas for surrender benefits a different less generous level of smoothing could be set. So payouts are closer to unsmoothed earned asset share.

Therefore the overall aim is for smoothing to be neutral over time, but the time period over which it is expected to be neutral for SVs is likely to be longer than for MV and DB.

 

[rlsrachaellouisesmith]

Yes I think you might be misinterpreting it. The cashflow addition approach allows more tailored addition of any surrender profits to policies, reflecting the experience within that cohort. (Although it could alternatively be done across a wider range of policies - as we explain in the second paragraph under the heading 'Cashflow addition to asset share' in Section 3.3 of Chapter 18.) The investment return method could be applied by just giving the same additional return to a wide range of policies / products, hence everyone gets the same amount - more broadbrush, less tailored, less reflective of the experience within particular cohorts. However, the latter could alternatively be done by cohort, as we explain in the second paragraph under the heading 'Addition via investment return'.

Thank you for confirming. So the choice of method is less about whether we are trying to apply to a wide range of policies or tailor it, more about systems/processes within the company.

For example, if they use deduction/addition from investment returns for other sources of credit/charge then they may choose this approach for consistency.

 

[rlsrachaellouisesmith]

Hi

When deductions are made for life cover from the AS, is the deduction the cost of life cover for benefit in excess of reserves or benefit in excess of AS. I expect it is benefit in excess of reserves, however, wanted to check.

Also, can a proportion of smoothing loss on surrenders/MV/death benefits be retained by estate if this is set out within the policy?

Thank you

 

[Lindsay Smitherman]

Therefore the overall aim is for smoothing to be neutral over time, but the time period over which it is expected to be neutral for SVs is likely to be longer than for MV and DB.

I'm not sure that holds, does it? For surrenders, if we are doing less smoothing then we might find that we are cost-neutral over a shorter time period? In the extreme, if we are doing no smoothing at all, then we are cost-neutral immediately

 

[Lindsay Smitherman]

When deductions are made for life cover from the AS, is the deduction the cost of life cover for benefit in excess of reserves or benefit in excess of AS. I expect it is benefit in excess of reserves, however, wanted to check.

It should be based on benefit in excess of asset share, because that is the amount that actually costs the life insurance company something and so would need to be recovered from the other WP p/hs.

 

[Lindsay Smitherman]

Also, can a proportion of smoothing loss on surrenders/MV/death benefits be retained by estate if this is set out within the policy?

Do you mean a proportion of smoothing profit being retained by the estate? That would rely on the company expecting to make such a profit over time, whereas normally such smoothing would be targeted to be cost-neutral. Having said that, as mentioned in Ch 18 Section 3.9, a charge might be deducted from asset shares in order to support the cost of providing smoothing, which is along the same lines (ie transferring something to the estate in relation to smoothing costs).

 

[rlsrachaellouisesmith]

Hi

A couple of questions here:

1)
If a smoothing account was used would it be correct to think of the balance of the smoothing account to be:
+estate contribution
+explicit smoothing charge
+(aggregate asset share - actual aggregate payouts)

2)
Is the shadow fund method for calculating Asset share equivalent to retrospective accumulation with product charges - from reading this does seem to the case. If this is true, why are they called out separately? If this is not true, what is the small difference that I have not noticed!

Thank you,

Rachael

 

[rlsrachaellouisesmith]

Do you mean a proportion of smoothing profit being retained by the estate? That would rely on the company expecting to make such a profit over time, whereas normally such smoothing would be targeted to be cost-neutral. Having said that, as mentioned in Ch 18 Section 3.9, a charge might be deducted from asset shares in order to support the cost of providing smoothing, which is along the same lines (ie transferring something to the estate in relation to smoothing costs).

I think what I meant was would a proportion of the losses be covered by the estate.

But since posting this I think I have got a bit of a better understanding, or at least I hope so. I have tried to summarise my understanding below.

If neutral cost of smoothing is not targeted then this means that there could be a smoothing profit or loss. Either way this smoothing profit or loss would be covered by the estate. Perhaps with an explicit charge for the cost of smoothing to AS which means a contribution would be made to the estate. Whether the aggregate charge was sufficient to cover smoothing costs or not is neither here nor there since the difference would fall to the estate.

I guess there is some consideration to p/h PRE and TCF. Need to ensure charge for cost of smoothing is explicitly called out. In terms of equity, if the WP fund closed then charges for smoothing costs might get returned to p/h who contributed them. In terms of equity if the fund is open then would we expect any 'smoothing profit' excess of aggregate charge for Cost of Smoothing over actual cost of smoothing to be returned to the remaining cohort that contributed?

Thank you

 

[rlsrachaellouisesmith]

I'm not sure that holds, does it? For surrenders, if we are doing less smoothing then we might find that we are cost-neutral over a shorter time period? In the extreme, if we are doing no smoothing at all, then we are cost-neutral immediately

Makes sense thank you!

 

[Lindsay Smitherman]

1)
If a smoothing account was used would it be correct to think of the balance of the smoothing account to be:
+estate contribution
+explicit smoothing charge
+(aggregate asset share - actual aggregate payouts)

Exactly how the smoothing account operates would be down to the insurer.

I'm not sure what you mean by '+ estate contribution' - do you mean as a sort of 'seed capital' to start the smoothing account off? That would not necessarily be required: there's nothing to say that the smoothing account cannot go negative - it's just a method of monitoring the cost of smoothing over time. As stated in the course notes, companies would generally aim to manage it so that it is close to zero.

Also bear in mind that not all companies would deduct a 'cost of smoothing' charge from asset shares.

And be careful with the final point: it would be more accurate to refer to adding the {difference between unsmoothed asset share and smoothed asset share for all policies becoming claims}. (Actual payouts might be greater than asset share because guarantees are biting, but that would not be part of the 'smoothing' account.)

 

[Lindsay Smitherman]

2)
Is the shadow fund method for calculating Asset share equivalent to retrospective accumulation with product charges - from reading this does seem to the case. If this is true, why are they called out separately? If this is not true, what is the small difference that I have not noticed!

The shadow fund approach is an example of a retrospective accumulation approach. The distinction lies within the mechanism being used: the parallel recording of the unit fund value for each policy on a daily basis (this is the 'shadow fund' - seen only by the insurer), but using actual investment performance rather than simply increasing unit prices by the regular bonus rate.

 

[Lindsay Smitherman]

I think what I meant was would a proportion of the losses be covered by the estate.

But since posting this I think I have got a bit of a better understanding, or at least I hope so. I have tried to summarise my understanding below.

If neutral cost of smoothing is not targeted then this means that there could be a smoothing profit or loss. Either way this smoothing profit or loss would be covered by the estate. Perhaps with an explicit charge for the cost of smoothing to AS which means a contribution would be made to the estate. Whether the aggregate charge was sufficient to cover smoothing costs or not is neither here nor there since the difference would fall to the estate.

I guess there is some consideration to p/h PRE and TCF. Need to ensure charge for cost of smoothing is explicitly called out. In terms of equity, if the WP fund closed then charges for smoothing costs might get returned to p/h who contributed them. In terms of equity if the fund is open then would we expect any 'smoothing profit' excess of aggregate charge for Cost of Smoothing over actual cost of smoothing to be returned to the remaining cohort that contributed?

Thank you

Yes - sounds like you have clicked it now!

In terms of your final question, that's unlikely. The accumulated charges would fall to the estate and remain there, ready to be used to support future smoothing costs.

The reason why charges taken from asset shares might be refunded to p/hs on WP fund closure is that the insurer is going to have to start distributing the estate to the WP p/hs who are in the fund at that point. Giving each p/h back what they have contributed to the estate feels like a reasonable first step in what is a complicated process of working out what a 'fair' distribution of the estate might be to each.

 

[rlsrachaellouisesmith]

Exactly how the smoothing account operates would be down to the insurer.

I'm not sure what you mean by '+ estate contribution' - do you mean as a sort of 'seed capital' to start the smoothing account off? That would not necessarily be required: there's nothing to say that the smoothing account cannot go negative - it's just a method of monitoring the cost of smoothing over time. As stated in the course notes, companies would generally aim to manage it so that it is close to zero.

That is exactly what I had in my head!

Thank you.