Australia’s Super Performance Test: The Regulation That Actually Worked

The performance test is the most successful piece of consumer financial regulation Australia has implemented in a generation, and it works through a mechanism most regulation does not: it makes underperformance fatal rather than merely embarrassing. A product that fails twice cannot accept new members, which means it cannot survive. The result has been a wave of closures, mergers and repricing that no amount of disclosure had previously achieved.

How does the test actually work?

By comparing what a product delivered to what its own asset allocation should have delivered. APRA constructs a benchmark portfolio using the product’s actual strategic asset allocation, applies representative index returns to each asset class, subtracts a representative fee, and compares the result to the product’s realised net return over the assessment period.

That design is what makes it hard to manipulate. A fund cannot claim to be a conservative product and then hold aggressive assets, because the benchmark is built from what it actually holds. It is measuring implementation skill and cost, not asset allocation choice, which is the right thing to measure in a default product.

The consequences escalate deliberately. A first failure requires written notification to members, which triggers outflows. A second consecutive failure closes the product to new members, ending its ability to grow. Because superannuation economics depend on scale, a closed product is a dying one, and trustees respond by fixing it, merging it or shutting it down.

What did the 2025 results show?

That the default market has largely cleaned itself up. APRA assessed 563 products in the 2025 test. All 52 MySuper products passed, including every major industry fund. All 374 non-platform trustee-directed products passed. The only failures were seven of 137 platform trustee-directed products, covering around A$1.02 billion of assets and approximately 8,500 member accounts.

That is a substantial improvement on 2024, when close to 20% of platform choice products failed. The pattern is consistent: platform products, accessed through advisers and offering wide investment menus, carry higher fees that drag on net returns, and the test measures net returns.

The pass rate should be read carefully. Passing means clearing a regulatory minimum, not delivering good performance. Analysis of platform products with ten-year histories has found that more than 40% still underperform their benchmarks materially, and the test only catches products that are measurably bad. Mediocre-but-passing products remain a large cost to members in aggregate.

What are the criticisms of the test?

Three substantive ones. The first is coverage: the test applies to more than 80% of accumulation-phase assets but only reaches a small fraction of platform products, and retirement-phase products are not tested at all. A retiree drawing a pension has no equivalent scorecard.

The second is benchmark rigidity. Because the benchmark uses listed index returns for each asset class, funds have an incentive to hug those indices rather than take genuine active positions, which may reduce long-run returns even as it reduces the risk of failing. Any single-metric test creates behaviour aimed at the metric.

The third is that annual measurement encourages short horizons in a system designed for forty-year outcomes. A trustee facing a possible failure has an incentive to avoid positions that are correct but slow to work, which is precisely the discipline long-horizon investors are supposed to have. There have been proposals to soften the test, and consumer advocates have argued strongly against them.

How has the test driven consolidation?

By making subscale funds untenable. A product that fails must explain itself to members and to APRA, and the cheapest path to a passing return is usually lower fees, which requires scale. Trustees that cannot reach that scale organically merge into someone who has it.

The result has been a long sequence of mergers — industry funds combining, corporate funds transferring members, and retail products closing entirely. Failed products are typically wound up with remaining balances moved into alternatives, which APRA has cited as evidence the test is working as intended.

The endpoint is a market of a small number of very large funds. That delivers lower fees and better bargaining power, and it concentrates systemic importance in a handful of institutions whose investment decisions move Australian markets. Whether the trade-off is favourable depends on how well those few institutions are governed — which is why regulatory attention has shifted from performance toward member services, valuation practice and administration.

What does this mean for trustees and employers?

For trustees, the test has become the central operating constraint. Investment strategy, fee structure and product design are all now evaluated partly by reference to their effect on the test result, and a product tracking toward failure triggers remediation long before the result is published. That is the intended behavioural effect, and it is powerful.

It has also reshaped fee negotiation with external managers. Because the benchmark includes a representative fee, every basis point of investment cost erodes the margin against the benchmark, and trustees have used the test as leverage to renegotiate manager fees across the industry. Managers who resisted repricing lost mandates.

For employers, the practical implication is narrower but real. Stapling means most employees arrive with an existing fund, so the default arrangement matters less than it once did. Where an employer still nominates a default, choosing a product with a consistent record of passing comfortably is a straightforward way to reduce the risk of a difficult conversation later.

How does the test compare with regulation elsewhere?

It is unusually direct. Most jurisdictions regulate retirement products through disclosure, requiring providers to publish fees and returns and trusting members to compare. Decades of evidence show that disclosure alone changes very little, because members do not read it and switching requires effort that inertia defeats.

Australia’s approach instead imposes a consequence that does not depend on member action. A failing product closes regardless of whether a single member reads the notification, which shifts the burden of protecting members from the member to the regulator and the trustee. That is a meaningful philosophical departure.

The United Kingdom’s value-for-money framework and similar initiatives elsewhere have moved in the same direction, and Australian regulators have presented the test internationally as a model. The main caution offered by observers is the one already visible domestically: a single quantitative test creates strong incentives to optimise for the test, and the design must be robust enough to survive that.

What should members actually do with the results?

Check three numbers rather than one. First, whether your product passed and by how much — APRA publishes product-level data showing the margin above or below benchmark, and a narrow pass is very different from a comfortable one. Second, the total fee on your specific investment option, not the fund’s headline administration fee.

Third, the net return over the longest period available on the option you actually hold. Many members are in a choice option rather than the MySuper default, and the fund’s advertised performance figures usually refer to the default. Those can differ substantially, and the difference compounds.

Then consider what you would lose by switching. Insurance held inside superannuation is often cheaper than equivalent retail cover and may not be replaceable on the same terms if your health has changed, so cancelling it by switching funds can be an expensive mistake. Exit fees are banned, so the cost of moving is time and insurance risk rather than money.

Where does the test go next?

Toward the retirement phase, most likely. The largest gap in the current framework is that pension products face no equivalent scorecard, at exactly the point where members have the most money at stake and the least capacity to recover from a bad outcome. Extending a comparable test to retirement products is the obvious next step and technically difficult, because outcomes depend on drawdown behaviour as well as investment return.

Coverage of platform products is the other unfinished item. Only a fraction of platform investment options are currently tested, and the segment where failures concentrate is also the segment least comprehensively assessed. Widening the net would be straightforward in principle and heavily contested by the providers affected.

There is also pressure in the opposite direction. Proposals to soften the test have been raised, on the argument that annual measurement discourages genuine long-term investment, and Treasury has consulted on the framework. Consumer advocates have pushed back hard, pointing out that more than 150 products have failed since 2021 and that the discipline is precisely what produced the current pass rates.

Frequently Asked Questions