Smoothing

We have two different types of smoothing: regular bonuses and payouts. in both cases we are trying to smooth out the volatility in investment returns.

First of all, consider regular bonuses on a unitised with-profits contract compared to a unit-linked contract (with the same underlying investments). The returns on the unit-linked fund each year might be:

10%, 15%, 3%, -5%, -5%, 10%, 15%.

We want the regular bonus rates to be much smoother, so we reduce them a little in the bad years, but only gradually. So regular bonuses might be:

4%, 4.5%, 4%, 3.5%, 3%, 3.5%, 4%.

Notice that the regular bonuses are not only smoother than the unit-linked returns, they are also lower as we are holding back surplus to build up a terminal bonus.

Secondly, lets consider the smoothing of payouts. Smoothing might be applied to death, maturity and sometimes even to surrender payouts. We normally set payouts with reference to asset shares. For example, the asset shares for 10 year endowments maturing in years X, X+1, X+2 etc might be:

100, 120, 130, 110, 90, 80.

We want to smooth the payouts so that the difference in payout from one yeat to the next isn't too large. So payouts might be:

100, 105, 110, 110, 105, 100.

Notice that the average payout over the 6 years is the same as the average asset share in this example. If the average payouts had been greater than the asset shares, we would have had to make up the difference by drawing on the estate - this could only work for a limited period, so in other time periods we might have to pay out less than the average asset share to build up the estate.

I hope these examples help. Good luck in the exam.

Mark

 

Hi
It won't necessarily be referring to terminal bonus. Payout smoothing can also apply on surrenders, where a terminal bonus may or may not apply.

Thanks
Em

 

Smoothing over time means averaging over time: smooth up when returns are poor, down when returns are good.
Deferring means declaring a lower bonus now (eg low RB), more later (eg high TB).

 

If the company is only using RB and does not use TB at all, then it is much harder to pay out exactly the asset share (this is what it means by "hit" the asset share). This is because TB can be set for homogenous groups of policies (eg split by duration in-force) specifically to meet the difference between asset share and guaranteed benefits awarded to date. RB is a more blunt tool - it just adds a given % each year, normally the same rate for all policies of a particular product. There is also more expectation of smoothing for RB rates than for TB rates, so that constraint also makes it harder to target the asset share.

 

There are two types of smoothing applied to with-profits business: smoothing of investment performance and smoothing over time. The former is a key feature of typical 'additions to benefits' WP business: policyholders are protected to some extent from poor investment performance but equally do not share fully in very strong investment performance, hence differentiating WP from unit-linked business.

Smoothing over time occurs because TB rates (which dictate the final payouts) are not changed on a daily basis. Policies maturing some time apart will receive the same TB rate, even though they have probably (to some extent) experienced different investment conditions.

If the company has a 'volatile' TB system, that means it changes its TB frequently. It can therefore change TB rates in line with actual investment performance, so that it can pay out close to asset share if it wishes to do so. However, it wouldn't be practical to change the TB rates on a daily basis, so precise asset shares would not be paid. Also, the company would still likely want to apply investment performance smoothing - hence also would not want to pay actual asset share.

If the TB system is more 'stable', that means that the company is not changing its TB rates very frequently. If, for example, it only changes TB rates at the end of each calendar year, a policy maturing in February will receive the same payout as an otherwise identical policy maturing in November, irrespective of the actual investment performance between February and November. If the performance is very strong over that period, the asset share of the policy maturing in November would likely be higher than that of the policy maturing in February, but they receive the same payout.

Hope that helps.