Geoff Warren (left) and David Bell

We’ve just released new research recommending the next essential step for development of a healthy retirement system following on from the introduction of the Retirement Income Covenant (RIC) and the first round of retirement income strategies (RIS). Our paper makes the case that RIS assessment frameworks should be established, and that this should start sooner rather than later. Funds, research houses, consultants and (potentially) regulators all have a role to play.

While our research targets super funds, we think there are some relevant applications for the advice sector.

Summarising our research

The principles-based RIC came into effect on 1 July this year. It provides significant flexibility for funds to develop RIS as appropriate to their membership. Meanwhile, differences exist across super funds in their memberships, the way they are approaching members in retirement and their capabilities. A large dispersion in the quality of RIS that will emerge from superannuation funds seems likely.

Assessment frameworks would help to provide some guardrails to complement the RIC, and corral the level of RIS dispersion. Designed well, they would frame what funds need to do while not curtailing flexibility and innovation.

However, assessing an RIS is challenging. Assessment needs to be mostly ex ante (forward looking) rather than focused on past performance. This is primarily because it is simply not feasible to wait decades to look back and assess what income has been delivered.

Furthermore, retirement is a multi-faceted problem. There are multiple objectives to balance (higher income vs. risk vs. access to capital), multiple sources of uncertainty (e.g. investment returns and mortality outcomes), multiple product types (account-based pensions, various lifetime income streams, and more) and multiple strategies for drawing income.

To top it off, the engagement piece is just as – if not more – important than how the assets are deployed. The RIC requires super funds to provide assistance to aid members to locate a suitable retirement solution. This also needs to be assessed.

We suggest two assessment approaches to be applied in concert.

The first is a qualitative checklist that evaluates whether funds are doing all that is needed to deliver a quality RIS. The checklist would consider a range of indicators to assess the suitability of the solution being offered, how effectively these are being managed, and the quality of the guidance being offered to members.

The second approach is quantitative modelling that simulates the range of possible retirement outcomes created by a fund’s solution. Assessment would entail gauging the improvement over a benchmark default solution such as an account-based pension coupled with minimum drawdown rates.

Applications to the advice community

While our research targets the specific challenges faced by the superannuation sector, there are some healthy reminders for the advice industry:

  1. The need for assessment of retirement strategies that are provided to members/clients resonates just as strongly for advisers. Best practice advice firms should have systems in place to ensure they are delivering quality outcomes to their clients. Advisers need to consider how this might be best be achieved given the size of their firms.
  2. The concept of a qualitative checklist seems relevant to advisers, although it would differ from that used by a super fund. For instance, a checklist might ask if recommended strategies draw on an appropriate set of product building blocks that address all relevant risks. Or how good are the processes for combining products and framing a drawdown strategy? Is there effective reporting and internal oversight of retirement plans? How well is the challenge of eliciting preferences and educating clients being met? And so on.
  3. Stochastic modelling should ideally be integrated into advice tools, likely as part of an advice firm’s tech stack. The aim is to assess how a strategy deals with the risks and trade-offs faced by clients through generating a range of possible outcomes. This level of assessment cannot be achieved with ‘deterministic’ models that generate a single pathway, and hence convey only the expected outcome rather than the possibility that it might not be achieved. We suspect many advisers are not currently supported by such tools.