How to choose and compare super funds

Superannuation is money your employer pays in on top of your wage, held in a fund you cannot draw on until you retire. The fund invests it, charges you for doing so, and the balance is yours. Which fund holds it, and how it is invested inside that fund, are the only two decisions in front of you, and together they are what choosing a super fund means.

If someone at work has handed you a form, it is the standard choice form, and it asks which fund your contributions should go to. You are allowed to leave it blank, and a set process decides what happens then. Your employer asks the ATO whether a fund is already attached to you from an earlier job and pays into that one if there is. That attachment is called stapling, and it means most people who have worked before are not starting from nothing. Contributions land in the employer's own default product only if you have never held a fund at all.

So the question is narrower than it first looks. It is rarely "which of the several hundred products in Australia suits me". It is "is the fund I would end up in anyway worth keeping, or is there a reason to nominate a different one". The account follows you between jobs from here either way, so the choice carries forward rather than coming up fresh at every new employer.

Working that out has an order to it, and it is close to the reverse of where most people start:

1. Can you join the fund? Some are open to anyone and some are restricted to a single employer or industry. This settles the question before any figure does.
2. Is it a MySuper product or a Choice one? MySuper products are the defaults, built to one legislated set of rules, and they are the ones that compare cleanly against each other. Choice products are menus of investment options you pick from, and they are reported differently. The section on the two below sets out the difference; if you have no view at all, MySuper is the kind the system puts you in by default.
3. Did it pass the annual performance test? A pass is a floor rather than a recommendation, but a product that fails twice running cannot accept new members, and the result is published for every MySuper product.
4. What are the total fees? Not the administration fee on its own. The total is what leaves the balance, and it is the only half of the comparison anyone can tell you in advance.
5. What has it returned over seven and ten years? A long record carries more than a good year, and the figure already has the fees taken out of it.
6. What does the Product Disclosure Statement say? Insurance is in there, it is often the largest deduction from a small balance, and it appears nowhere in a fee comparison.

The rest of this guide is the detail behind steps three to five: what each published figure measures, which ones can be compared across funds, and which ones cannot.

Superannuation is unusual among the products on this site: there is no daily feed and no advertised rate. What exists instead is an annual publication from the Australian Prudential Regulation Authority, which collects the same fields from every fund on the same basis and publishes them together.

The most important part of that basis is the balance. APRA’s headline figures report fees and returns as they would apply to a $50,000 account, for every product, whatever its members actually hold. It also publishes fees at other balances, but $50,000 is the common measuring stick. That matters because a large part of most administration fees is a flat dollar amount, and a flat fee is a very different proposition on $15,000 than on $400,000. Fixing the balance is what stops a fund looking cheap purely because its members are wealthy.

The consequence for a reader is worth stating plainly. A fee figure on a comparison table is not a quote for your account. If your balance is well under $50,000 the percentage you pay is probably higher than the published one, and if it is well over, lower. The published figure is a like-for-like measuring stick, not a bill.

Fees are the half of a super comparison that is knowable in advance. A fund cannot tell you what its investments will earn next year, but it can tell you exactly what it will charge you to hold the account, and it has to publish it.

Three charges make up the figure APRA reports. The administration fee covers running the fund and your member account. The investment fee covers managing the money, and it varies far more between funds than the administration fee does, because it depends on how much of the portfolio is actively managed or held in unlisted assets. Transaction costs cover buying and selling within the portfolio. Added together they are the total fee, and it is the total that comes out of your balance.

This is why a comparison sorted on the administration fee alone can mislead. A fund can run a very low admin fee and a high investment fee, or the reverse, and a member of either pays the sum. Total fees show what leaves the balance, and the split shows where the cost is going.

One structural point is easy to miss. Because part of the administration fee is usually a flat dollar amount, the percentage a small account pays is higher than the published one. A member with $8,000 spread across an old account can be paying several times the headline rate on that balance, which is the arithmetic behind most of what is written about combining accounts.

Insurance premiums are separate again. They are deducted from the balance like a fee, they are often larger than the administration fee, and they are not part of the figure APRA publishes. A fee comparison does not see them, and neither does a table built from one.

A return figure looks like the answer to the whole question, and it is the figure most people go to first. It rewards a little care, because APRA publishes more than one kind and they measure different things.

A net return is what a $50,000 balance actually grew by after investment fees, administration fees and tax. It is the closest thing to a member’s own experience, and because fees are already taken out of it, a net return is not a separate topic from the fee comparison. It has the fees inside it.

A net investment return is what the investments earned after investment fees and tax but before administration fees. It measures the investing on its own, separate from the cost of running the account, which is useful when you want to see whether a cheap product is cheap because it charges less or because it invests differently. It always sits above the net return for the same product.

The periods matter as much as the measure. A one year return says more about which asset classes had a good year than about the fund, since a product weighted to shares will lead a rising market and trail a falling one without anything about the fund having changed. Seven and ten year figures cover more than one market cycle, which is why long-term returns carry the most weight. A blank in the ten year column usually means the product has not existed for ten years, not that it returned nothing.

The other thing a return figure carries is risk, which the number itself does not show. A product invested heavily in growth assets, meaning shares, property and infrastructure, will generally produce a wider spread of outcomes in both directions than one weighted to cash and bonds. Two products with the same ten year return can have taken very different paths to get there. That is why APRA also publishes the investment mix, and why the performance test compares each product against a benchmark built from its own mix rather than against the field.

Each year APRA runs a test on every MySuper product, and on trustee-directed Choice products, comparing its investment returns against a benchmark built from that product’s own investment mix, and its administration fees against the median for similar products. Because the benchmark follows the product’s own strategy, a conservative product is not marked down for behaving conservatively. The test asks whether the product did as well as its own strategy should have, after fees.

The result is published as a pass or a fail. A product that fails must write to its members and say so, and a product that fails two years running cannot accept new members. That closure is the real consequence of a fail.

A pass is a floor rather than a verdict. It tells you the product cleared a benchmark set from its own mix; it does not rank it against the products that also passed, and it says nothing about whether the strategy suits any particular member. Where APRA attaches its own footnote to a result, it publishes an asterisk alongside it, and a comparison should show the result exactly as published rather than tidying it away.

MySuper products are the defaults. If an employee never chooses a fund, this is the kind of product their employer’s contributions land in, and every MySuper product has to meet the same legislated rules. That uniformity is what makes them properly comparable with one another: one product, one set of figures, one performance test result.

A Choice product is anything else, usually a menu of investment options the member picks from. It has no single net return to publish, because two members of the same product invested in different options had different years. Figures are published per option instead, which is why a Choice product appears as a range rather than a number, and why the widest range on a table usually belongs to the product with the most options rather than to the most volatile one.

Two further traps sit inside the Choice data. Some Choice options are outside the performance test altogether, so a blank result there means untested rather than failed. And APRA publishes returns for options offered through an investment platform on a different basis from the rest, so the two sets cannot be compared with each other, and a table that sorts them together produces a ranking that means nothing.

One last wrinkle inside MySuper. A lifecycle product moves members between investments as they age, and publishes a separate set of fees and returns for each age stage, sometimes dozens of them. There is no single figure that is true for the whole product, so a lifecycle product shows as a range and the stage that applies to you depends on your date of birth. A single strategy product invests every member the same way and has one set of figures.

A comparison table is a good way to see what products charge and what they have earned. It is a poor way to see everything else, and these are the gaps:

None of the following appears in the published figures:

• Insurance, which is often the largest deduction from a small balance. Cover, premiums and the terms of a claim vary enormously between funds and none of it is in the fee figure.
• Whether you can join. Some funds are open to anyone and some are restricted to a particular employer or industry, which decides the question before any figure does.
• The investment mix behind the return, and how much movement you would have lived through to earn it.
• Service, insurance claims handling and the retirement products a fund offers when you get there, none of which is a reportable number.
• Your own circumstances, including your age, your balance, your tax position and what else you hold. Two people reading the same table are not looking at the same decision.
• Contributions and tax, which are set by law rather than by the fund and are the same wherever the account sits.

The documents that do cover the gaps are the Product Disclosure Statement and, for Choice products, the Target Market Determination, which funds publish alongside them. The PDS sets out the fees in full, the insurance and the terms; the TMD sets out who the product was designed for. They are the primary source, and a comparison table is a way of narrowing down which ones to read.

If a decision turns on something the figures do not carry, particularly insurance or tax, that is the point at which licensed financial advice earns its cost. CompareThat compares published figures and does not advise.