Core reading question 16.4

Hi Himanshu

In Question 16.4 in the notes we have two unit-linked contracts. The first contract has no guarantees - the policyholder can use the lump sum to buy an annuity at whatever price is being charged at the time. The second contract is the same except that it offers an option to buy the annuity at a guaranteed rate - the policyholder is charged an extra 1% of the fund each year to pay for the guarantee.

Part (ii) asks us to describe a model that could assess whether the extra 1% charge is appropriate. The charge will be appropriate if the profitability of the two contracts is similar. So the solution sets out first of all how to assess the profitability of the first contract, and then does the same for the more complex second contract. It then compares the result of the two to highlight any situations where the second contract is less profitable (which would indicate that the 1% charge was unsuitable).

There are several steps to follow in describing a model. These steps are not unique to pricing unit-linked guarantees. So in general we have:

1) state the purpose of the model
2) determine whether the model is stochastic or deterministic
3) choose the parameters in the model (and which will be stochastic)
4) determine the inputs, these may be model points
5) gather data to set assumptions, probability distributions
6) consider correlations between parameters and any dynamic links
7) build the model, describing the cashflows
8) run the model, this may include stochastic simulations or deterministic sensitivity tests
9) check the goodness of fit and revisit the earlier steps to improve the model if necessary
10) communicate the output to key stakeholders who will make a decision.

I hope this helps to explain the solution.

Best wishes

Mark