C21 - Analysis Of Experience DONE

Question 1

Reasons for monitoring actual experience

Answer 1

• To check if method and assumptions adopted for financing benefits continue to be appropriate
• What changes should be made in order to achieve desired level of prudence
• Gives a realistic estimate of cost of benefit provision
• May lead a sponsor to change it’s policies eg discretionary increases

Question 2

Data required to monitor actual statistical experience of a DB scheme

Answer 2

• Tabulated record of all events that have occurred at each age (or any other significant factor affecting the frequency of the event eg gender)
• Number of individuals exposed to those events at each age
• Assumptions adopted and any potentially relevant standard tables

Depending on volume of data a multi way analysis may be carried out

Question 3

Data to monitor the economic experience of a scheme

Answer 3

Relevant current and past economic data corresponding to the timing of regular checks on funding position
Eg for salary growth need to collect details of salaries grouped by factors closely correlated to salary growth

Question 4

When may there be insufficient data for a credible analysis of experience to be performed?

Answer 4

If you are monitoring:

• small schemes
• infrequent and volatile event eg death before retirement
• data is subdivided into too many homogenous subgroups

Question 5

Analysis of mortality experience

Answer 5

• Calculate for each age band the number of deaths divided by the number exposed to the risk of death
• Compare this with the assumptions adopted and standard tables to determine whether there is a significant difference or if those assumptions or tables are appropriate
• If appropriate, carry out more statistical analysis to determine the standard deviation of actual experience

Question 6

Which groups of people should not usually be included in an analysis of salary experience?

Answer 6

• New joiners since last valuation
• Leavers
• People who joined and left between the valuation dates

Question 7

Analysis of salary experience

Answer 7

Produce a table with the columns:

• Age x nearest at current valuation
• Salaries at previous val of members age x
• salaries at current val of members age x
• actual increase in salaries
• expected increase in salaries
• Actual / expected increase

Stable pattern of differences at different ages may show a correlation between salary growth and age. To analyse promotional/age related increases separately, identify the actual global cost of living increases and assumed promotional scale

Question 8

Crude analysis of salary growth

Answer 8

• Divide this times average salary by last time and take the third root (for inter valuation period)
• Dangerous as it can be severely distorted by new entrants and retirements
• Gives an initial impression of salary experience and can be refined by restricting the comparison to the average salary of members who were present at both valuation dates

Question 9

Factors to consider before using the results of any analysis

Answer 9

• Was the period typical?
• Is the experience likely to be representative of future experience eg abnormal events or significant random fluctuations
• If there are changes over time, is there any reason why past trends carry on in the future?

Question 10

Give the financial assumptions that often lead to surpluses/deficits

Answer 10

• Investment return
• Salary increases
• Pension increases
• Post retirement mortality

Question 11

Reasons a period may not be typical or representative include:

Answer 11

• trends
• cycles
• abnormal events
• random fluctuations
• data errors or changes in recording
• heterogeneity in the group
• changes in the balance of homogenous groups