Despite being a sector which has always been required to maintain data on its members (i.e. customers), superannuation funds have been slow to embrace data analytics. Part of the reason for this is that administration systems designed primarily to record basic member data have typically not been flexible enough to readily provide more valuable information for management.
With the scale of many funds now allowing improvement in technology, data on fund members has never been more accessible. Funds can now extract and analyse information on their members’ transactions and selections easily and relatively cheaply. One analysis that can be performed is an assessment of a member’s risk tolerance.
The default investment option drives the retirement outcomes for most fund members. When setting or reviewing a fund’s default option, funds need to understand the attitude and ability to take risk of their membership. Without such an understanding, funds may be taking too much risk or not enough.
Understanding the risk tolerance profile of a fund’s membership is fundamental to addressing questions like:
- Should we consider age-based lifecycle option as our default?
- Should our default be the same pre- and post-retirement?
- Should we consider different default options for different member sub-groups?
- What communication strategies should we be using for our different member sub-groups?
While there are a number of ways in which risk tolerance could be assessed (e.g. by conducting a survey), a data-driven approach is appealing as it makes use of the very rich information most funds already hold which can provide valuable insights into their members’ risk tolerance.
There are four key drivers or factors of risk tolerance for an individual. 1 Net wealth. 2 Life influences like gender, age and health. 3 Human capital like education and job security. 4 Level of governance, such as time, interest and expertise.
For each member, each of the above factors can be assessed, or scored by reference to basic membership data a fund holds. For example, as superannuation is the largest asset most individuals own outside the family home, the account balance is a good proxy for an individual’s net wealth. By aggregating the scores derived for each of the four factors, we can then assess the overall risk tolerance profile of a fund’s membership.
By then conducting and comparing the same analysis for different segments within the membership (for example, by employer, contribution category, location, and so on), we can also gain insight into the diversity and segmentation of risk tolerance across the fund. Further, we can use a similar modeling approach to assess other behavioural traits of the membership, such as propensity of members to make an investment choice as against being a true defaulter.
Now let’s consider assessing the adequacy of members’ retirement savings.
To develop their communications, marketing and product design strategies, funds need to understand how well their members are progressing towards an adequate retirement. While forecasting tools are useful for individual members and their advisors, funds need whole-of fund information so they can address questions like:
- How are our members progressing towards achieving reasonable retirement outcomes?
- Are there pockets of the fund’s membership that have fallen behind and might need special attention?
- How do our default settings and members’ choices affect their potential retirement outcomes?
A data-based modelling approach can generate useful insights. Starting with each member’s current account balance and their (inferred) weekly or annual earnings, individual balances are projected to retirement age, converted to an equivalent income, and the member’s estimated government age pension entitlement added. The resultant income is divided by an adequate income benchmark – based on either the ASFA standards, or the individual’s pre-retirement earnings – to derive a ratio (which we term the Adequacy in Retirement Index, or ARI) indicating the percentage of adequacy income the member is currently targeting.
When aggregated, the ARIs show a picture of the adequacy of the whole funds’ membership broken down by age, employer or other demographics, which can be tracked over time. And, perhaps most usefully, funds can consider “what if” scenarios by varying the various inputs and assumptions – such as the impact on retirement adequacy of higher contribution levels, varied investment return levels and earlier and later retirement age.
While not exhaustive, these examples show how funds can generate information that adds real value and insight to their strategic decision-making. With technology barriers receding, in future a wider range of member data will only become more accessible. But those funds that can harness that data most effectively will be best placed to succeed in aligning their product offerings and default settings with the needs of their members.
Nick Callil is the head of retirement income solutions, Australia for Towers Watson
