AOS - economic variances

[Mbotha]

1. The matching adjustment surplus relating to asset transactions is described within the economic variances section of the AOS chapter. Wouldn't this actually be captured under economic assumption changes though? Any change in matching adjustment due to changes in backing assets would only come through in the next valuation through an assumption change to the discount rate used.
2. Similarly for changes in asset mix (the core reading example given speaks about WP business and the impact on volatility assumptions if a stochastic approach is used). Wouldn't this also fall under economic assumption changes?

 

[Viki2010]

I think that the way the classification works in AoS is that the "change in Economic or Non-Economic assumptions" sections include changes when the company resets the rates - i.e. not the actual differences A vs E.

 

[Lindsay Smitherman]

Yes, the "economic variances" item relates to impacts arising from economic (or asset-based) events that have happened over the year of analysis and the "changes in economic assumptions" item relates to impacts arising from a changed view of what is going to happen after the end of that period.

As stated in the Core Reading, it is not always easy to separate these aspects: "Changes in economic assumptions may be shown separately or may be combined with the above (economic variances), due to the close relationship between actual economic outcome over the year and the best estimate future economic assumptions, particularly in relation to yields and bond value movements."

A change in asset mix (which underlies both of your examples) is an actual management action which has taken place during the year, rather than being a change in the company's assumptions as to what will happen in future, so probably sits better within "economic variances" in that respect. However, as stated in the Core Reading, changes in asset mix "may be separately identified in the analysis" - ie have their own separate line.

Hope that helps.

 

[Naimin Patel]

Hey,
When looking at April 15 qu 2)iv) it states under investment return variances that this would no longer be needed Under VIF, is this because assets are modelled separately?

 

[Lindsay Smitherman]

No - it is because there is zero VIF for this situation (as stated earlier in the question).

 

[Naimin Patel]

I understand that, I think I'm just getting confused as to how actual experience has no impact with reference to VIF, ie the actual investment return as opposed to the expected. But I see they mention this as being deduced from measuring change in assets to the change in bel.

 

[gruhaa]

Yes, the "economic variances" item relates to impacts arising from economic (or asset-based) events that have happened over the year of analysis and the "changes in economic assumptions" item relates to impacts arising from a changed view of what is going to happen after the end of that period.

As stated in the Core Reading, it is not always easy to separate these aspects: "Changes in economic assumptions may be shown separately or may be combined with the above (economic variances), due to the close relationship between actual economic outcome over the year and the best estimate future economic assumptions, particularly in relation to yields and bond value movements."

A change in asset mix (which underlies both of your examples) is an actual management action which has taken place during the year, rather than being a change in the company's assumptions as to what will happen in future, so probably sits better within "economic variances" in that respect. However, as stated in the Core Reading, changes in asset mix "may be separately identified in the analysis" - ie have their own separate line.

Hope that helps.

Hi Lindsay
I have a different view. When management changes its assets mix(having both the mentioned effects) shouldn't it change the company views on future in terms of returns and volatility due to change in assets position/structure during the analysis period?
I also have question related to mismatching profit.
To be honest, I couldnt understand why mismatching profit(also can you explain it with an numerical example) is getting considered when we allowed for actual investment return variance(difference in asset value and liability value post investment return) and also a change in surplus position due to change in asset yield under 'economic chnage' bucket.? I mean any surplus due to change in asset yield/investment return assumption shouldn't be due to mismatching between asset n liability(not adequately immunized against change in yield) ? And can you also help me understand the last line of this 'Mismatching profit' paragraph-'impact of the yields on the actual backing assets rather than swap rates used for the BEL'? What impact it being talked here

Thanks

 

[Lindsay Smitherman]

I have a different view. When management changes its assets mix(having both the mentioned effects) shouldn't it change the company views on future in terms of returns and volatility due to change in assets position/structure during the analysis period?

As I mentioned above, different companies will categorise a change in asset mix differently within the analysis of surplus - so different views are fine! The original question was why the impact of a change in asset mix might be categorised as an economic variance rather than change in assumption, hence the response.

You are correct that a change in asset mix would change the volatilities that are modelled - but this can be considered to be different from changing the volatility assumptions. A change in volatility assumption could be changing your view from future equity market volatility being 20% to assuming that future equity market volatility will be 22%, say, eg because you have performed better analysis. Changing asset mix is a management action which doesn't involve changing your estimates of what the future might be (the average volatility modelled will be different due to having different asset class weightings, which are known amounts, not due to making different assumptions about the volatilities for each asset class) - hence it could be categorised separately from assumption changes.

Bear in mind also that the expected investment return would not change as a result of a change in asset mix, assuming that we are talking about an analysis of surplus on a Solvency II basis. This is because Solvency II liabilities are determined on a risk-neutral market-consistent basis, so future expected investment returns are set at the risk-free rate, irrespective of actual assets held.

 

[gruhaa]

As I mentioned above, different companies will categorise a change in asset mix differently within the analysis of surplus - so different views are fine! The original question was why the impact of a change in asset mix might be categorised as an economic variance rather than change in assumption, hence the response.

You are correct that a change in asset mix would change the volatilities that are modelled - but this can be considered to be different from changing the volatility assumptions. A change in volatility assumption could be changing your view from future equity market volatility being 20% to assuming that future equity market volatility will be 22%, say, eg because you have performed better analysis. Changing asset mix is a management action which doesn't involve changing your estimates of what the future might be (the average volatility modelled will be different due to having different asset class weightings, which are known amounts, not due to making different assumptions about the volatilities for each asset class) - hence it could be categorised separately from assumption changes.

Bear in mind also that the expected investment return would not change as a result of a change in asset mix, assuming that we are talking about an analysis of surplus on a Solvency II basis. This is because Solvency II liabilities are determined on a risk-neutral market-consistent basis, so future expected investment returns are set at the risk-free rate, irrespective of actual assets held.

Thanks a lot Lindsay for such an insightful reply.
Any help you can give me on mismatching profit question I aksed in the same post.

 

[Lindsay Smitherman]

To be honest, I couldnt understand why mismatching profit(also can you explain it with an numerical example) is getting considered when we allowed for actual investment return variance(difference in asset value and liability value post investment return) and also a change in surplus position due to change in asset yield under 'economic chnage' bucket.? I mean any surplus due to change in asset yield/investment return assumption shouldn't be due to mismatching between asset n liability(not adequately immunized against change in yield) ?
Thanks

Hi - the Core Reading states that 'the economic variances item will include mismatch profits (or losses) where the assets backing the liabilities do not move perfectly in line with the liabilities following a change in economic conditions.'.

Let's say that the assets and liabilities are perfectly matched. In that case, if market values or yields of assets change, there will be an exactly offsetting change in liabilities. Hence no surplus will arise. However, if they are not perfectly matched (as will normally be the case) then any change in value or yield on assets will result in an element of surplus. That is what this Core Reading statement is referring to.

It is important not to consider surplus arising from economic variances as being the impact on the assets and the impact on the liabilities as two separate isolated elements. We should be considering them together.

Hope that helps.

 

[Lindsay Smitherman]

And can you also help me understand the last line of this 'Mismatching profit' paragraph-'impact of the yields on the actual backing assets rather than swap rates used for the BEL'? What impact it being talked here

Remember that the risk-free rates used to discount the BEL are based on swap rates for the UK calibration. Even if liabilities are closely matched using government bonds, say, then the assets and liabilities will not move exactly in line with each when the yield curve changes, if yields on government bonds and yields on swaps do not move in exactly the same way as each other. Hope that makes more sense now.