Dynamic management actions & policyholder behaviour

[Edward chong]

Hi,
I would like to ask that, apart from with-profits policies, is it common for other business lines to have dynamic management actions & dynamic policyholder behaviour included in Solvency II internal models in computing technical provisions & solvency capital requirements?

Can you please suggest some examples for applicable business lines & their relevant management actions & policyholder behaviour?

What are the specific requirements to include these management actions into internal models?

Thank you.

 

[Lindsay Smitherman]

Hi - one example would be for a unit-linked whole life assurance product where there is a guaranteed minimum surrender value payable at the ten year anniversary, say. An example of dynamic policyholder behaviour that could be modelled would be low surrender rates in the period running up to the ten year policy anniversary if the guarantee looks like it will be in-the-money at that date under a particular simulation or scenario (policyholders being more likely to hang on to obtain the guarantee) and high surrender rates at the ten year anniversary if the guarantee "bites".

An example of a dynamic management action might be where a unit-linked policy has a variable charge: the model might allow for the charge to be increased under a scenario where expenses are high.

 

[Mandla]

An example of a dynamic management action might be where a unit-linked policy has a variable charge: the model might allow for the charge to be increased under a scenario where expenses are high.[/QUOTE]

Might the point at which management can increase charges not be defined as the contract boundary? And there discretion can be used on whether or not to model cashflows beyond the contract boundary when calculating BEL liabilities?

 

[Lindsay Smitherman]

Might the point at which management can increase charges not be defined as the contract boundary?

Possibly, but not necessarily. It is only a contract boundary if the charges can be changed in "such a way that they fully reflect the risks". This is a difficult area of interpretation in practice (which is why the Core Reading doesn't go into this, and so you wouldn't be expected to know about the details). For example, if the variable charges were capped in some way, it could be argued that they cannot "fully reflect the risks" and so would not constitute a contract boundary.

 

[Lindsay Smitherman]

And there discretion can be used on whether or not to model cashflows beyond the contract boundary when calculating BEL liabilities?

There is no discretion on the use of contract boundaries: if there is a boundary then no future premiums (or cashflows arising from those premiums) can be taken into account.