SA3: A Selection of course notes questions

[Joe Warren]

Hi,

After going through the SA3 course notes I have a number of questions I would appreciate if I could get help answering:

1) Mutuals & Tax: Chapter 3, page 11 states 'Underwriting losses and profits arising from mutual trading may be exempt from tax. It may be that no tax relief is given for expenses which are part of the mutual trade'. I was wondering what expense items would receive no tax relief, whereas for a proprietary insurer they would?

2) Deciding on the Discount Rate: Chapter 9, page 9 states 'Usually, the insurer does not need to deduct tax from the discount rate because claims are tax deductible.' If we are discounting by a rate based upon by the expected return of the assets that back the liabilities, then wouldn't we need to account for tax in the rate as the return from the assets will be taxed?

3) Discounting & Free Reserves: Chapter 9, page 11 states (wrt discounting) 'We are effectively just moving funds within the balance sheet from the (now smaller) technical reserves to the (correspondingly larger) free reserves.' I understand why technical provisions reduce, but why do the free reserves get larger?

4) SII and future profit: Chapter 9, page 15 states 'Solvency II allows the recognition of future profit to a certain extent'. This is stated as a difference to IFRS17; I was wondering is the recognition of future profit in relation to the discounting that is performed under SII, but discounting is also permitted under IFRS17, so is this recognition not permitted under both standards?

5) Premium Allocation Approach: Chapter 9, page 15 states under the PAA that the liability for unexpired risk can be taken to be the unearned premiums, and that a CSM is not required. Under the BBA, the aim of the CSM is to recognise losses on onerous contracts immediately, but not to allow recognition of future unearned profits. Under the PAA approach, is this still a condition that must be met? If the insurer has priced the policy appropriately, then by assuming liabilities equal unearned premium then are we assuming we will be making a profit, and as a result would we need to hold an amount equal to the profit component of the unearned premium and release it as the profits earns to ensure that it earns over the remaining period of exposure?

6) Novation: Chapter 14, page 14 states that the impacts of outwards reinsurance will be considered in determining the payment to be made to the new insurer with a Novation. Is this to say that if a contract that is to be transferred was previously affected by a reinsurance arrangement under the original insurer, would the reinsurance arrangement transfer and continue to cover the contract, and thus reduce the payment size that is required to be made?

Many Thanks,

Joe

 

[Ppan13]

1) Mutuals & Tax: Chapter 3, page 11 states 'Underwriting losses and profits arising from mutual trading may be exempt from tax. It may be that no tax relief is given for expenses which are part of the mutual trade'. I was wondering what expense items would receive no tax relief, whereas for a proprietary insurer they would?

Claims handling expenses.

 

[Ppan13]

Hi,

3) Discounting & Free Reserves: Chapter 9, page 11 states (wrt discounting) 'We are effectively just moving funds within the balance sheet from the (now smaller) technical reserves to the (correspondingly larger) free reserves.' I understand why technical provisions reduce, but why do the free reserves get larger?

Joe

The free reserves is the balancing item item on the balance sheet. Since the total assets actually being invested is not being changed, if the technical reserves are lowered by discounting, the allocation to free reserves has to increase to balance it out.

 

[Ppan13]

Hi,

6) Novation: Chapter 14, page 14 states that the impacts of outwards reinsurance will be considered in determining the payment to be made to the new insurer with a Novation. Is this to say that if a contract that is to be transferred was previously affected by a reinsurance arrangement under the original insurer, would the reinsurance arrangement transfer and continue to cover the contract, and thus reduce the payment size that is required to be made?

Many Thanks,
Joe

I might be wrong but I don't think it's saying the outwards reinsurance HAS to be transferred in all novations, as opposed to 'should be considered'. It would be logical to include the outwards reinsurance in the negotiation of the novation in order to continue RI protection on broadly similar terms, and in the case of Part VII transfers the court usually does transfer the outwards reinsurance contracts attributable to the underlying business.

I'm not sure including the outwards reinsurance in the novation would always reduce the payment size; you'd expect the net liabilities to be less than the gross liabilities but presumably the old insurer would want something in return for 'selling' the OWRI asset to the new insurer (given the former had ceded premium to the reinsurer, and reinsurance can be expensive)

 

[Ppan13]

2) Deciding on the Discount Rate: Chapter 9, page 9 states 'Usually, the insurer does not need to deduct tax from the discount rate because claims are tax deductible.' If we are discounting by a rate based upon by the expected return of the assets that back the liabilities, then wouldn't we need to account for tax in the rate as the return from the assets will be taxed?

I don’t know the answer to this one. I see the logic of what you’re saying, more so than whatever logic the notes are trying to make with regards to the tax-deductible characteristic of claims . At the same time, when I look online, I see this paper making the claim that, “Discount rates will reflect the nature of insurance contract liabilities under IFRS 17 rather than the nature of the assets held” being listed as an “improvement in IFRS 17 compared with IFRS 4”, which seems to go contrary to the argument you’ve made (though I don’t know the reasoning for the quoted statement.) See table 1 in https://www.bis.org/fsi/publ/insights26.pdf

There’s also this old paper on Discounting in GI by an IoA working party (from 1987!) which says (paragraph 6.3, p10) , “There has been some theoretical argument over whether the rate of discount should be on a gross or net of tax basis. The working party has concluded, through worked examples, through logic and through parallels with appraised values, that a gross rate is appropriate where discounted provisions are used for tax purposes. Accounts drawn up on this basis would produce a zero profit in each year of a run off situation where claims and interest payments are accurately known and where the asset income matches the claim outgo.” https://www.actuaries.org.uk/system/files/documents/pdf/discounting.pdf

Again, they don’t explain more about the logic they talk of or provide the worked examples, so I’m left with a question mark in my mind about this one for now!

Hopefully someone else reading the forum will have clearer insight. Or maybe ask on one of the Life forums, since discounting has been more traditionally applied there.

 

[Joe Warren]

I might be wrong but I don't think it's saying the outwards reinsurance HAS to be transferred in all novations, as opposed to 'should be considered'. It would be logical to include the outwards reinsurance in the negotiation of the novation in order to continue RI protection on broadly similar terms, and in the case of Part VII transfers the court usually does transfer the outwards reinsurance contracts attributable to the underlying business.

I'm not sure including the outwards reinsurance in the novation would always reduce the payment size; you'd expect the net liabilities to be less than the gross liabilities but presumably the old insurer would want something in return for 'selling' the OWRI asset to the new insurer (given the former had ceded premium to the reinsurer, and reinsurance can be expensive)

Thanks - makes sense

 

[Joe Warren]

Thanks to Ppan13 again for his responses - it is much appreciated.

If anyone viewing this thread has further thoughts on Q2, and any thoughts at all on Q4 and Q5 it would be much appreciated.

 

[Ppan13]

You're welcome, Joe.

Re: Q4, { "SII and future profit: Chapter 9, page 15 states 'Solvency II allows the recognition of future profit to a certain extent'. This is stated as a difference to IFRS17; I was wondering is the recognition of future profit in relation to the discounting that is performed under SII, but discounting is also permitted under IFRS17, so is this recognition not permitted under both standards?" }

This is what I had in mind regarding the comparison of treatment of future profit between SII and IFRS17:

Future contracts from existing or potential customers are not included in the Solvency II balance sheet, but you could have premiums yet to be received on existing in-force business (as opposed to future business), and therefore potentially future profits. Gross future premium (and claims cashflows) for policies not yet incepted by the valuation date but already forming part of contractual obligations will also form part of the premium provision. More generally, when considering the future cashflows in SII, when future income is greater than future outgo then expected future profit can be recognised.

Under IFRS17 however, the Contractual Service Margin doesn’t allow recognition of future unearned profit (CSM is set to equal the profit, whilst profit is set to zero in the accounts). The CSM only “unwinds” over time to release the profit to the P&L as it is earned.

 

[Busy_Bee4422]

I don’t know the answer to this one. I see the logic of what you’re saying, more so than whatever logic the notes are trying to make with regards to the tax-deductible characteristic of claims . At the same time, when I look online, I see this paper making the claim that, “Discount rates will reflect the nature of insurance contract liabilities under IFRS 17 rather than the nature of the assets held” being listed as an “improvement in IFRS 17 compared with IFRS 4”, which seems to go contrary to the argument you’ve made (though I don’t know the reasoning for the quoted statement.) See table 1 in https://www.bis.org/fsi/publ/insights26.pdf

There’s also this old paper on Discounting in GI by an IoA working party (from 1987!) which says (paragraph 6.3, p10) , “There has been some theoretical argument over whether the rate of discount should be on a gross or net of tax basis. The working party has concluded, through worked examples, through logic and through parallels with appraised values, that a gross rate is appropriate where discounted provisions are used for tax purposes. Accounts drawn up on this basis would produce a zero profit in each year of a run off situation where claims and interest payments are accurately known and where the asset income matches the claim outgo.” https://www.actuaries.org.uk/system/files/documents/pdf/discounting.pdf

Again, they don’t explain more about the logic they talk of or provide the worked examples, so I’m left with a question mark in my mind about this one for now!

Hopefully someone else reading the forum will have clearer insight. Or maybe ask on one of the Life forums, since discounting has been more traditionally applied there.

My take is something like this.
We are doing all this because we are doing a market consistent valuation so what we want is market values of liabilities. The BE of liabilities is defined as the value that another person would be willing to take them over in an arms-length transaction which is essentially the market value of liabilities. In choosing the yield curve (the assets held may or may not match the liabilities) to use to discount we want to pick one where it has similar characteristics as the liabilities, particularly the risk and tax characteristics. Regarding the tax status of assets, they get taxed via capital and income gains but liabilities are tax exempt. It, therefore, doesn't appear suitable to use the net tax discount rate.

 

[Busy_Bee4422]

Hi,

5) Premium Allocation Approach: Chapter 9, page 15 states under the PAA that the liability for unexpired risk can be taken to be the unearned premiums, and that a CSM is not required. Under the BBA, the aim of the CSM is to recognise losses on onerous contracts immediately, but not to allow recognition of future unearned profits. Under the PAA approach, is this still a condition that must be met? If the insurer has priced the policy appropriately, then by assuming liabilities equal unearned premium then are we assuming we will be making a profit, and as a result would we need to hold an amount equal to the profit component of the unearned premium and release it as the profits earns to ensure that it earns over the remaining period of exposure?

Joe

Hi

The PAA is calculating the liability for remaining coverage as the unearned premium reserve whilst the BBA splits this up further into its components ie CSM, the expected value of future cash flows, and the risk margin. Using the PAA still meets the requirement that future profit is not recognised.

In the event that a policy is onerous then a loss component needs to included on top of the unearned premium reserve. There are details on how to do this in greater detail if you look at sample accounts by the audit firms.

 

[Joe Warren]

The free reserves is the balancing item item on the balance sheet. Since the total assets actually being invested is not being changed, if the technical reserves are lowered by discounting, the allocation to free reserves has to increase to balance it out.

Hi Ppan13 - I've just been thinking about this more. I understand what you are saying, and if you look at a balance sheet if the asset value doesn't change then it makes sense for the free reserve value to increase as the liabilities reduce.

However, let us say that our liabilities are undiscounted and we perfectly match our assets to cover our liabilities. Then the undiscounted value of both assets and liabilities would be the same. If the value of the assets in the balance sheet can be based upon undiscounted future income, if we were then to discount our liabilities, would we not then discount the asset future proceeds as well and as a result have a smaller asset value as well, thus both reducing?

Or, if the value of an asset is taken to reflect the discounted value of future proceeds, would we not need to hold more of the asset to cover the undiscounted liabilities? Then if we were to discount the liabilities, we would need to hold less of the assets that are valued based upon discounted future proceeds in order to match the now smaller liabilities, and thus hold a lower amount of assets?

 

[Ppan13]

Hi Ppan13 - I've just been thinking about this more. I understand what you are saying, and if you look at a balance sheet if the asset value doesn't change then it makes sense for the free reserve value to increase as the liabilities reduce.

However, let us say that our liabilities are undiscounted and we perfectly match our assets to cover our liabilities. Then the undiscounted value of both assets and liabilities would be the same. If the value of the assets in the balance sheet can be based upon undiscounted future income, if we were then to discount our liabilities, would we not then discount the asset future proceeds as well and as a result have a smaller asset value as well, thus both reducing?

Or, if the value of an asset is taken to reflect the discounted value of future proceeds, would we not need to hold more of the asset to cover the undiscounted liabilities? Then if we were to discount the liabilities, we would need to hold less of the assets that are valued based upon discounted future proceeds in order to match the now smaller liabilities, and thus hold a lower amount of assets?

You could look at it that way, i.e. hold the free reserves as constant and reduce your investments in line with the reduction technical reserves. In the original scenario in the notes, they were holding the assets constant, so in their P&L, the total investment returns (on insurance funds + on free reserves) was unchanged. In the scenario of reducing your assets instead, the investment returns will be reduced in the P&L, which would to some degree offset the impact of higher retained profit caused by the larger outstanding c/f (on a discounted basis).

The main point of the scenario in the notes was (i think) to demonstrate that if you simply introduce discounting into the accounts (without actually changing anything else, e.g. how much total assets you invest), by itself there is no actual change to the company's cashflows, because it's just a notional change in accounting for reserves (shifting some allocation from the technical reserves to the free reserves).

If in reality you change your level of investments based on a choice of discounting, that has potential implications on solvency margins (e.g if you discount too heavily, leading to underinvestment compared to a best estimate or non-discounting case), so a regulator might seek to place limits on the permissibility / level of discounting.

 

[Joe Warren]

Thanks Ppan13 & Busy_Bee4422 for your replies!

 

[Ppan13]

Thanks Ppan13 & Busy_Bee4422 for your replies!

You’re welcome , and thanks for asking those questions, as they really made me think and test my own understanding of the themes (and what I’m missing!)

 

[Busy_Bee4422]

Thanks Ppan13 & Busy_Bee4422 for your replies!

You are welcome.