ifrs 17 and csm

[dimitris13]

Hi ,

is csm part of the liability in the BS ?
or is it like a "separate calculatio" showing its effect on p&l ?

thanks

 

[Em Francis]

Hi ,

is csm part of the liability in the BS ?
or is it like a "separate calculatio" showing its effect on p&l ?

thanks

Hi Dimitris
For SA2 purposes, I would think about the CSM as being a component of the liability in the balance sheet. And to think of the balance sheet being made up of:

• the fulfilment cashflows and the time value of money
• the risk adjustment (RA)
• the CSM.

Thanks
Em

 

[dimitris13]

Hi Em ,
many thanks on this.
just one more question with regards to csm. in general when we say release of [something] eg reserves we mean: reserves(t+1)-reserves(t).
So i was expecting that the release of the csm would be similarly csm(t+1)-csm(t) that goes to p&l as profit (or loss).
instead i see sth like: -(csm(t)+csm(t)*i) * amortisation_factor. are these equivalent ?

 

[mugono]

It’s probably useful to make a distinction between 1. the actual timing of cash flows (ie when reserves and the CSM would be released into profit and loss) and 2. assumptions someone may make about the cash flows timing to simplify the calculation.

Your definition of release of reserves and change in CSM contains an implicit assumption about the timing of the cash flows (ie falls within 2 above).

It’s always helpful to think about what, if any, assumptions (implicit or otherwise) a particular result relies. The implication is that there is unlikely to be a single correct approach. The important thing is to be consistent throughout the calculation.

From Em’s liability definition, CSM is a component of the reserves. I’d therefore expect the release in reserves to include components over and above those being attributed to the change in CSM.

Hope that helps.

 

[Mateusz]

Hi Em ,
many thanks on this.
just one more question with regards to csm. in general when we say release of [something] eg reserves we mean: reserves(t+1)-reserves(t).
So i was expecting that the release of the csm would be similarly csm(t+1)-csm(t) that goes to p&l as profit (or loss).
instead i see sth like: -(csm(t)+csm(t)*i) * amortisation_factor. are these equivalent ?

For a General Model (GM), a simplified roll-forward of the CSM for a sample group of contracts would like the following:

CSM(T+1) = CSM(T)
(+) Effect of NB
(+) interest accretion (*)
(+/-) Changes in BEL+RA related to future service
(-) CSM amortisation (*)

The items followed by (*) are those which will be reflected in the P&L (CSM accretion as 'insurance finance income or expenses' and CSM amortisation as a component of 'insurance revenue').

The impact of the other components on the total insurance liabilities is zero, and this is why these are not released to the P&L. As an example, let's say we changed operating assumptions for the valuation of BEL + RA. Assume the impact of this change is X.

Now:

BEL + RA impact: + X
CSM impact: - X (changes in BEL+RA related to future service)
Total liability change = 0 = Total P&L impact

Similar to the effect of NB: at inception, the CSM is set to ensure no net gain/loss. So for a profitable group the NB CSM is, by definiton, set to -(NB BEL + NB RA), which gives the total liability (and P&L) impact of 0.

The above is just to illustrate the ideas, in practice things get more messy than this. Examples:
- if an increase in BEL/RA related to future service is greater than the CSM balance, any excess will flow to the P&L as a loss because the CSM cannot be lower than 0
- if you issue NB which is onerous (i.e. BEL + RA > 0 as inception), you'll report a loss immediately, and so there will be a non-zero effect of NB on the P&L.

M.