A Review of Retirement Savings Investment Behaviours: Theory and Evidence


June 2013

This study provides a review of individual retirement savings decisions within the Australian superannuation system, and analyses when and why investors make (or don’t make) changes to their superannuation savings. Using the Mercer database, our preliminary analysis of member investment activity over the 2003−12 period suggests that some key demographic factors (gender, age, balance, income) are associated with investor behaviour. Other factors such as financial literacy and internet access also appear to explain investment activity. The most compelling finding is the lack of activity. The majority of members do not make changes even over considerable periods of time.*

Superannuation has become the second largest asset of Australian workers after the family home. The superannuation funds that members are enrolled in by their employer, to meet mandatory superannuation contributions obligations, are primarily defined contribution (DC) funds.

DC funds place the individual member at the centre of responsibility for investment decisions regarding these contributions and, with time, their accumulated savings balance. A defined contribution (DC) plan is a type of retirement plan in which the amount of the employer's annual contribution is specified and, unlike a defined benefit plan (DB), only employer contributions to the account are guaranteed, not the future benefits.

Active choices by members of superannuation funds were assumed to become key forces within the system driving competition and improving overall efficiencies. This view has, however, been acknowledged as ‘optimistic’ (see Australian Government 2010 Intergenerational Report). The thrust of the more recent wave of regulations, such as MySuper, challenges this assumption more directly and has reoriented the choice architecture with the member as a reluctant decision maker at the forefront. This reflects concern about the interest/engagement of members in the task at hand as well as their competency to deal with the sophistication of products offered.

In considering investment choices, the dominant framework within investment finance has at its core rational, utility maximising, risk-averse investors who aim to maximise returns while minimising risk. Alternative models have been proposed to consider investment choices, largely drawing on behavioural models from psychology.

We examine individual retirement savings decisions, and provide an overview of when and why investors make (or don’t make) changes to their superannuation savings. We start by presenting the Australian setting and reviewing the levels of choice that members of superannuation funds have available. We also review the available literature to examine the influences on choice suggested by theory and reported empirically, paying particular attention to the influences of financial literacy and demographics in these decisions.

Next, we turn to potential behavioural explanations for superannuation choices as well as the choices involved in making retirement savings. The review is not exhaustive but attempts to highlight the principal areas of relevance to the Australian setting.

Our report concludes with a preliminary analysis of member investment activity using a new database provided by Mercer Australia, a major wealth management company. We present selected preliminary analysis of a new database of member investment behaviour constructed from members of the Mercer Super Trust, Corporate Division.

The Mercer Database includes approximately 200 employer sub-plans with approximately 400,000 employees. These members can be tracked between 2003 and 2012 as they enter, receive employer superannuation contributions and make investment strategy changes. The data includes both active and exited members.

Three different types of investment activity are observable:

  • a contributions investment change (CIC) which changes the investment strategy for future contributions
  • a balance investment change (BIC) which changes the investment strategy of the accumulated balance
  • a combination of a BIC and CIC at the same time.

These are treated as separate investment change events for purpose of the analysis here.


The Australian retirement income system continues to mature as it enters its twenties. As with any young adult, significant changes are being encountered. Major structural changes to the institutional environment, through SuperStream, as well as the investment product environment through MySuper will influence the retirement income well-being of Australian workers.

From this review, it is clear that our knowledge of investor behaviour and their decisions regarding investment savings and changes in their retirement savings can be improved. There appear to be some key demographic factors (gender, age, balance and income) associated with investment behaviour. We also find that other factors such as financial literacy and internet access can explain investment activity.

The most compelling finding is the lack of activity. The majority of members do not make changes even over considerable periods of time. Our understanding of why this is the case is improving but further reconciliation with available empirical facts is needed. Ideally this would utilise both revealed behaviour, as from the Mercer database, and attitudinal data collected via survey from the member when the investment is (and is not) undertaken.

* We thank the essential contributions of Rohan Fletcher and Mikhail Tupitsyn in database management and preliminary analysis of the Mercer data. We also gratefully acknowledge the provision of data by Mercer Australia.