The first 47 pages of the consultation paper explore the various pros and cons of different tests. The last two pages tacked on the end eventually turn to the crucial question of what the consequences of failing the test should be.

This highlights a flaw in how this issue is being approached. Any form of APRA performance test is merely a means to an end. You need to know what the end is before you can start designing the means. As the Cheshire Cat told Alice, if you don’t know where you’re going, any road will get you there.

Until now, the end has been a curious one. The end is not what you might think it to be.

For starters, the current APRA performance test is not about rewarding good performance. It punishes poor performance. This can be contrasted with other flash-in-the-pan thought bubbles. The now-defunct ‘best-in-show’ regime would have anointed outstanding performers through inclusion in a blessed top ten list of funds that could receive default contributions. All funds other than the best ten would have missed out equally.

There is a protective and paternalistic element too.  Again, all is not what it seems. It is not quite right to say that the current regime protects Australians from languishing in poorly performing funds. Even when a poorly performing product is forced to close, everybody who has already invested in that product remains and can continue investing more money in that same product (funds merging out of existence was an outworking of the regime, but not actually part of the regime). It is more accurate to say that the current regime is a bollard that prevents extra people from entering poorly performing products.

Clarity is needed on what the endgame is going to be in future. Let’s assume the goal remains some sort of a safety net: to do something about products that perform badly, rather than rewarding top performers. Immediately, there is no incentive (from the endgame) to chase exceptional performance and every reason to minimise the risk of poor performance.

The next question is what the consequence should be for poor performance. Should the safety net only be there for the next person who comes along? Or should something be done about the people who are already invested in the product?

This is where consumer sovereignty and libertarian attitudes can enter the equation, especially when talking about choice products. The current regime stops people from choosing a product – and stops superannuation funds from even offering a product - if the product has failed the test two years in a row. If this is the goal, and if you value sovereignty and business liberty, perhaps the test should use metrics that identify products that have such undeniably dire poor performance that the choice and freedom to offer the product deserve to be taken away. Sovereignty and liberties should not be lost because of hair triggers and test metrics which are known to have flaws, especially if a variety of test metrics are used and the results go both ways. 

It is worth pointing out a key difference between choice products and MySuper products. The government arguably has a broader mandate to set tougher standards for MySuper products. These are the products that have the privilege of receiving the compulsory contributions that are made for Australians who have not chosen a superannuation product. Community expectations are higher when it comes to safeguarding forced savings by those who have not taken action to safeguard those savings themselves. Hair triggers and the rough justice of a flawed test might have been justifiable in the past when the context was MySuper products only.

Choice products have been chosen by people who are free to choose again to exit, having been established by trustees and promoters who may have invested significant amounts to establish those products. It is fair to ask whether the test for choice products should be designed to identify undeniably dire poor performance with no false positives and no after-the-siren complaints of the outcome being swayed by flawed test design – if closure is the consequence of failure.

On the other hand, there is less pressure to have a perfect test if the consequence of failing the test were to be something less than closure (for example, mandatory disclosure to choice members to ensure they are fully informed of the situation). In saying that, much turns on the form of any compulsory disclosure, because punitive disclosures could have a reputational and cashflow impact akin to closing the product to new members.

For funds that have MySuper products, most of their members and most of their assets are invested in the MySuper product. If the MySuper product fails the test, some take this as a reflection on overall governance and capability, rightly or wrongly. Because of the centrality of the MySuper products to some fund’s ongoing existence, the closure of the MySuper product can precipitate the closure of the whole fund. For the above reason, some may say that outcome is warranted.

The situation is different in the choice fund context. A fund will have numerous choice products. Many might pass the APRA performance test and, in some years, one choice product might fail. Remember that all the products offered by a fund are all oversighted by the same trustee board and the same investment committee and possibly even managed by similar investment managers. Outside of the MySuper context, failing the test does not give rise to the same question mark over governance and capability. The same individuals will likely have successfully managed numerous other choice products that did pass the test. Fairness dictates that this be borne in mind when deciding on the appropriate consequences of failing the test. The forced closure of one choice product could have reputational and cash flow impacts on other choice products that successfully passed the test – including for the members in those other products.

Adopting an approach that utilises multiple metrics, and that focuses on products that unambiguously fail all of the metrics before forcing the product to close, might be accepted as fairer.

The metrics would still need to be carefully chosen though, having regard to the context. The government has suggested that metrics from APRA’s heatmaps could be commandeered. What might normally be a ‘bad’ metric could be a ‘good’ or ‘mitigating’ metric for poorly performing choice products.

Take APRA’s sustainability metrics, for example. Normally, it is a ‘good’ metric if the number of accounts are increasing and cash flows are strong. However, this could be a red flag if the product is a poorly performing choice product. In contrast, if a poorly performing product is seeing a smaller number of accounts and cash outflows as members exit, this could be evidence that the choice system is working – that members are voting with their feet. If so, perhaps there might be less of a need for regulatory intervention, until scale and liquidity become a concern.