September 2020 Paper 2 Q2(ii)

[KMER94]

Just going through the examiners report for paper 2 for last September. I sat this paper and struggled particularly with this question and can see from the examiners report other people struggled (with the people scoring on average less than 25%). I'm wondering how this question should be approached when there's not much on accounting data or revenue accounts in the course notes? (Mind, it's been years since I did CT2..)

The question is as follows:

Background:
A large proprietary life insurance company writes a range of products in its home market. It has implemented the principles of treating customers fairly throughout all aspects of its business.

The company has been making losses on its portfolio of annuities in payment for the last 5 years. The Board of the company feel that a significant factor in these losses is due to poor policy data for its annuity business. The pricing actuary of the company has also expressed concerns that the rate of future improvements in longevity will be higher than was assumed in the original pricing of the existing annuity business.

The Board of the company are also under pressure from the company’s shareholders to start selling new business in developing markets overseas. The Board are therefore considering a potential opportunity in a large overseas Country X. To date, the only life insurance policies that have been allowed to be sold in Country X have been through the state nationalised insurer. The state nationalised insurer offers only a very limited range of products, and significant parts of the population have not purchased any life insurance products. Proprietary life insurance companies have just started to be allowed to compete with the state nationalised insurer. The Company is considering whether to start selling life insurance business in Country X.

The insurance regulator in Country X is considering how it might assess whether customers have been treated fairly, by surveying customers when they receive the payouts on their products, and then asking how happy the customers are with their payout.

(ii) Discuss how the company’s accounting data could be used to assess the policy valuation data of the annuity business. [13]
​

I would have never have gotten the three headings used in the answer, which were:

I – compare annuity pa with regular/annuity payments/claims in revenue account
II – compare new annuity reserves with new single premiums in revenue account
III - compare terminating annuity reserves with lump sum claims in revenue account​

Any tips?

 

[Lindsay Smitherman]

Don't worry - this is a very challenging question so don't be too disheartened if you have struggled to generate ideas for it. Also don't assume that splitting the solution up in this way is the only way to tackle it. In our ASET Solutions for that question, we have used a different approach to answering it (which still would gain the marks): thinking about the advantages and limitations of using accounting data to check policyholder data.

In terms of specific checks that could be used, consider the cashflows that would appear in a revenue account (premiums, claims etc) and then which of these could be used to check against policy valuation data. For example, the accounts will include the actual regular annuity benefits paid - is this consistent with the in-force annuity benefit amounts in the policy data? The accounts will include the actual new annuity premiums received during the period - is this consistent with the reserves set up for annuities sold during that period?

 

[singingmyblues]

Hello, may I ask what would the policy data include? eg
* policyholder info (not estimated, but collected) - DOB, sex
* financial values that involve actuarial judgement - reserves at individual policyholder value, PV(future annuities) etc?

 

[Lindsay Smitherman]

Yes to policyholder info.
Not necessarily the actuarial calculations though - these are more what is calculated using actuarial data, not the data itself.
But policy data would also include information about the policy, such as the amount of sum assured / benefit, the frequency and level of premium, the start and maturity dates.

 

[singingmyblues]

Yes to policyholder info.
Not necessarily the actuarial calculations though - these are more what is calculated using actuarial data, not the data itself.
But policy data would also include information about the policy, such as the amount of sum assured / benefit, the frequency and level of premium, the start and maturity dates.

thank you for the reply! As the mark scheme talked about comparing actual new annuity premiums received during the period (from accounting data) with reserves set up for annuities sold during that period, I was thinking the latter is from policy valuation data. Basically, are individual reserved contained in policy valuation data?

 

[Lindsay Smitherman]

As I said above, not necessarily.
But bear in mind that the 'policy valuation data' (ie the data used to perform the policy valuations) will be used to calculate those individual reserves. So if comparing the reserves against other items indicates that there might be a problem with the reserves, then this could be pointing to a problem with the underlying valuation data used - so that is also part of the potential assessment of data quality.
So, it doesn't actually really matter whether or not you interpret 'policy valuation data' as including individual reserves or not - the same sorts of points can be made (because issues with reserves could imply issues with the underlying valuation data either way). So in the exam don't worry too much about things like this, as it is usually the case that a good enough answer can be produced however you interpret the information. Hope that's sufficiently reassuring!