Stranded balances, fee pressure, and APRA’s retirement scorecard

This fortnight the retirement income policy agenda is moving on multiple fronts at once, and the connections between them are hard to miss.

The Actuaries Institute has published its most concrete proposal yet for converting the super sector into a true retirement income system. UniSuper has fired a competitive shot in pension-phase pricing that the rest of the large fund sector will notice. And APRA has released the consultation that will, for the first time, make retirement-phase outcomes publicly measurable. Read on.

The Actuaries Institute has released a dialogue paper co-authored by Nick Callil, (previously the Head of Retirement Solutions at Willis Towers Watson), and David Knox AM (recently retired Senior Partner at Mercer and one of Australia’s foremost retirement thought leaders), making the case that Australia has a well-respected superannuation savings system, but not an actual retirement income system. The paper proposes a practical package of measures to overcome this decumulation disconnect.

The paper estimates that more than 1.5 million Australians aged 65 and over still hold superannuation balances worth approximately $326 billion in accumulation phase, despite being eligible to transfer to tax-free pension phase. These 'stranded balances' are costing members more than $2 billion in additional tax each year. Inertia, complexity and the absence of any effective default mechanism are identified as the core culprits.

The solution proposed by Callil and Knox, branded 'MyIncome', is a three-legged package designed to progressively shift members toward pension phase, without removing individual flexibility before age 75.

The first leg requires APRA-regulated funds to offer a pre-set, fund-designed account-based pension to all accumulation members from age 65, with the offer repeated annually and made as frictionless as possible: no forms, no complex decisions, just an acceptance.

The second leg mandates that any balance still in accumulation phase at age 75 must be compulsorily transferred to pension phase, covering both APRA-regulated funds and SMSFs. The third requires APRA-regulated funds to begin capturing and validating bank account details and member identity from age 60, removing a key friction point before the first offer is made at 65.

Under the proposals, each fund would set its own MyIncome drawdown and investment parameters; broadly analogous to the freedom funds currently exercise in designing their MySuper products. The paper explicitly excludes a longevity component from the initial package, not because the authors are unsympathetic to lifetime income, but because requiring any 'fact find' to determine lifetime income suitability would undermine the core premise: a smooth, near-automatic transfer available to all members in the first instance.

The MyIncome proposal is the one of the more operationally concrete default decumulation frameworks to emerge from an industry or professional body to date, and it deserves serious consideration. The stranded balance data alone is a damning indictment of the status quo.

The paper's decision to exclude a longevity component is pragmatically sound. The system first needs to normalise regular pension drawings over lump sum withdrawals and capital hoarding. Add lifetime-income complexity before that norm is established and you risk defeat. As we wrote recently, there is still much more that can be done with account-based pension design on issues like sequencing risk, before a longevity sleeve adds further risk mitigation.

As to the design of MyIncome, further investigation into the behavioural drivers and preferences of members who deliberately stay in accumulation-phase past 65 is warranted. Are some members purposely choosing to remain in accumulation? Do they see the tax foregone (beyond 60) as an option premium for the flexibility of not ‘eating capital’ prematurely? Do these members vary by sociographic circumstance?

Another difficult issue for MyIncome will be fund operational readiness. Capturing and validating bank details from age 60, designing fund-specific default drawdown and investment parameters, managing anti-hawking compliance and communicating the offer clearly to members who have never engaged with their fund, will be far from trivial at scale.

Operational complexity notwithstanding, this new Actuaries Institute paper is a reminder that in super, Australia has an present a very effective long-term savings program struggling to make income the outcome.

UniSuper has announced a 50% reduction in the asset-based administration fee on its Flexi Pension products, one of the most significant unilateral fee cuts by a major Australian superannuation fund in the retirement phase in recent memory. The reduction applies to both the Retirement Phase and Transition to Retirement Flexi Pension products, and takes effect from 1 July 2026.

UniSuper already occupies a strong position in the pension fee league tables, consistently ranking in the top quartile on administration costs across Chant West's annual Pension Fund Fee Survey. Halving its asset-based admin fee to 0.08% will sharpen that position considerably.

UniSuper modelling suggests that the average Flexi Pension member will save around $495 per year, based on the average Flexi Pension (Retirement Phase and Beneficiary Income Stream) account balance as at 28 February 2026. That matters because UniSuper, as a profit-to-member fund primarily serving the university and higher education sector, has a member base with above-average balances and longer-than-average retirement horizons. Pension-phase fee competitiveness will increasingly be consequential for that constituency, particularly as larger-balance members consider for-profit platforms (often with the assistance of advice).

The significance of this announcement extends beyond its dollar impact on UniSuper members. It is a direct challenge to the rest of the larger fund sector, at a moment when the retirement phase is under unprecedented regulatory and public scrutiny.

BFID requires trustees to act in the best financial interests of beneficiaries, so when a peer fund halves its retirement product admin fees (while maintaining its investment performance and service standards) pricing status quo becomes harder to sustain.

UniSuper’s move can be read as a statement about which way the competitive winds are blowing in member retention and retirement readiness.

Trustees who have not recently benchmarked their pension phase fee structures should be doing so before the transparency of the Retirement Reporting Framework (see below) arrives.

APRA has released a consultation paper setting out its proposed approach to implementing the Retirement Reporting Framework (RRF), one of several retirement-phase reform measures announced by Government in November 2024 and finalised in February 2026.

The RRF is intended to support consistent, transparent monitoring of the outcomes delivered to members in retirement, and to help trustees measure their progress against the retirement income strategies required by the Retirement Income Covenant.

APRA's role is to collect the data annually from late 2027, with first publication slated for 2028. The data will also be shared with ASIC, Treasury and other relevant government agencies, and used jointly by APRA and ASIC to track progress on Covenant implementation.

The proposed data collection is built around three indicators and three metrics. The indicators operate at an RSE level and assess the offerings funds make available to members: whether they provide drawdown options beyond the minimum rate (Indicator 1), whether they offer access to lifetime income products (Indicator 2), and whether they provide or facilitate access to personal financial advice (Indicator 3).

The metrics go deeper, capturing member behaviour: take-up of retirement products including lifetime income products (Metric 1), actual drawdown rates and benefit payment values (Metric 2), and balance utilisation at account closure including death (Metric 3). All metric data will be segmented by member cohort attributes; age, gender, account balance band, retirement status indicators, and product holdings.

To implement the Framework, APRA is proposing to introduce a new reporting standard (SRS 611.1 Retirement Member Profile), amend the existing SRS 607.0 (RSE Business Model), and revoke SRS 610.0 (Membership Profile) to reduce duplication.

The revocation of SRS 610.0 is worth noting; APRA has effectively determined that most of its remaining data will either be superseded by SRS 611.1, or is of limited continuing value.

For those wishing to make a written submission on the Retirement Reporting Framework as proposed, the due date is 3 June 2026.