See below for the better comparison using tax drag and index return data over the same period as the APRA data
I read today that the Australian tax office had put up a superannuation (retirement) fund comparison tool, using data collected by APRA. Apparently it's a new government initiative. It is a step in the right direction in that the tool let's you easily compare total dollar fees for a portfolio of any given size (and your age, which I don't fully understand). The tool does not report any kind of risk measure, nor any kind of risk-adjusted return, but it does give six year return numbers, data current as at 31 March 2021. It looks like the ATO plans to update the tool every quarter. Importantly the data is for the default (the so-called MySuper) fund for each superannuation fund complex. No default fund I've ever read about indexes, but these are the funds most people use because they don't make a conscious and informed choice.
When there is data, I feel an uncontrollable urge to graph. Below is plotted the 6 year average fund return versus total dollar cost for a $100,000 portfolio of someone aged 42. The portfolio size was chosen to match the comparison portfolio size in our wiki page about Australian super (and conveniently you can divide dollars by 10 to get cost in basis points). The age is just midway between 18 and 65. 57 funds out of 80 total made the list. The 23 funds not making this list were labeled restricted, i.e. not open to the general public, so I restricted the analysis to funds that anyone could chose. Eight funds that did not have return numbers were also excluded, leaving 49 funds in the analysis. Most of the funds are industry (non-profit) funds, but a handfull are for-profit funds. The black dots in the figure are the data for the 49 funds. The median fee and the median 6 year return were respectively 1% and 6.6% per year; these median values are labeled by the gray open circle.
While the data itself is interesting, I find it even more interesting to imagine what all these super fund members in default funds could have earned and could have paid in fees if the default funds were comprised of a few low cost index funds. This is one of the failures of the comparison tool: no attempt is made to benchmark against an objective standard or against the policy portfolio of the fund. First the fees.
The vertical red line represents the fee level for the $100,000 portfolio if using the index fund options at SunSuper or Aware super (formerly FirstState). And these are not the lowest cost indexed options around; the indexed balanced option at HOSTPlus, for example, costs around $140 per year. The red line indicates that you could pay a third or less than the median cost of the default funds by chosing a super complex that let's you index. As Bogleheads, we know these extra costs compound: high costs rob us of the dollars we need for a comfortable and secure retirement. The graph makes clear that the default fund fails on cost compared to what you could pay by chosing the indexed option.
But what about returns? There's an old nostum that better returns are worth paying more for. Let's see what the data for Australian super has said over the last 6 years. The first thing to notice is that there is zero correlation between fees and returns for the default funds (black dots). Whether you paid $500 or $1500 per $100,000 invested did not predict what kind of returns you got. As I expected a negative correlation, I was surprised at zero correlation, but I suspect it's a quirk of the short 6 year time-frame.
More interesting is the question, what returns could these funds and their members have gotten with an indexed global balanced fund? I compute the return to a 70:30 stock:bond portfolio using the real returns reported by wholesale (institutional) index funds at Vanguard AU. And keep in mind that this is not a theoretical exercise: every one of these super funds could have been invested this way via Vanguard or Blackrock or any number of other index fund managers. I've chosen to construct a portfolio as
- 60% Vanguard International Shares Index Fund, 5 year return 12.88%
- 10% Vanguard Australian Shares Index Fund, 5 year return 10.11%
- 20% Vanguard International Fixed Interest Index Fund (Hedged), 5 year return 2.52%
- 10% Vanguard Australian Fixed Interest Index Fund, 5 year return 3.10%
The indexed portfolio that any of the super funds could have used returned 9.5% a year and is represented by the horizontal red line in the figure. This return is higher by 1.4% a year than even the best performing super fund. The large magenta dot in the figure at the intersection of the vertical (costs) and horizontal (returns) red lines marks the location of a real, and markedly superior, portfolio that any of these default funds could have used; and, that any individual using the non-default, index investment options at real super funds could have made for themselves. A caveat to this is tax. I believe the APRA return numbers are net of tax, whereas the constructed portfolio had no tax taken out. I don't know how much super funds pay typically in tax; however, even knocking off 15% of the indexed portfolio return for tax only reduces the index portfolio return to the level of the best performing default funds.
The bottom line, which experienced Bogleheads already know, is that costs matter---a lot. Even more so because performance is ephemeral, while costs are eternal. (I'm pretty sure that's a Bogle paraphrase.)
Did I play fair in making comparisons? The indexed portfolio has more in ex-AU stocks than I suspect is typical in an Australian super fund, at a time when ex-AU did better than AU stocks; although, that's also a choice---and a risk---to be markedly different from the global stock market index. But, as a bit of a sensitivity analysis, let's mark down the ex-AU stock index fund to have the same return as the AU stock index fund, equivilent to assuming zero investment in ex-AU stocks, which is also atypical for super funds. Now that's the darker magenta dot, and the indexed portfolio returns 7.9% a year, a hair below the best performing default funds, but still comfortably above the median return. Even if we knock off 15% more for tax, our indexed portfolio is still a bit above the median return (6.7% vs 6.6%). And the indexed portfolio is guarenteed to earn market returns, whereas the actively managed default funds take on the extra risk of falling short of market returns.
The data says it's just a fact, and a shame for the retirement security of the fund members, that the default high-cost active management funds have fallen woefully short of what the members could have earned at lower cost in a global balanced indexed portfolio.