Returns for superannuation funds over the last 10 years have been excellent, in fact, Bridgewater Associates recently produced a chart showing excess returns (i.e. returns above the cash rate) were the second highest for any 10-year period out of the past 50.
So, if things aren’t broken, they don’t need fixing — right? Wrong.
The Bridgewater Associates research outlines how a simple, low-cost passive equity and bond portfolio (what we refer to as a reference portfolio) has performed over different 10-year periods. Happily, for many superannuation fund members, an unforeseen consequence of the push towards simple, liquid and low-cost, portfolios has been truly remarkable returns. However, this outcome is just that – unforeseen and remarkable.
Our concern is that this outcome is having a strong influence on government policy and the broader direction of the superannuation industry, with little regard for the likelihood of experiencing the same remarkable set of outcomes in the future.
The last 10 years has been an exceptionally good decade for returns on equities and assets in general. If we consider that the biggest driver of this has been a significant fall in real interest rates, related to a huge increase in global liquidity driven by central bank activities, it is very difficult to see how this can be repeated.
But let’s look at where superannuation was in December 2009 – almost half way between the introduction of the Choice and MySuper regimes, and just beginning to understand the cause and effects of the GFC. The GFC resulted in a heightened level of risk aversion, and saw investors seek low complexity and liquidity. At the same time, the Choice/MySuper regimes led to a huge focus on peer-relative outcomes and, in particular, a strong competitive drive to lower costs. While we are very strong advocates for removing excessive costs from the system, this focus has produced some unintended consequences.
The desire to reduce investment management fees has driven many funds towards the use of passive strategies, and the experience of the last 10 years will have helped justify these decisions. Indeed, the Productivity Commission’s ‘best in class’ suggestions and APRA’s ‘member outcomes’ approach are dangerously close to creating a strong belief that the returns produced by large, low-cost funds in the recent past are repeatable, and a reflection of the skill of those funds.
However, we believe it’s time to provide a counter to this. The use of past performance has its limitations, and so investors should always go beyond looking just at performance in making their decisions - they also need to take a robust and more balanced approach to their analysis.
Let’s first consider the problem with just looking at past performance as the basis for making a decision. From a statistical perspective; when there are large allocations to passive equities in our investment portfolio, we are opening ourselves to a wide range of future outcomes, because equities are very volatile assets.
By definition, if a fund’s investment strategy is open to a wide range of possible outcomes, members can experience very strong performance, but equally they can also experience some very weak outcomes. The chart below gives a sense of this. Looking at any 10-year period for a simple equity and bond reference portfolio since 2000, you can see the wide dispersion in realised outcomes, from a cumulative (10 year) excess return over cash of +6.4% p.a. to -1.5% p.a.
The purple line in the graph represents performance over the last 10 years to June 2019 (which is cash +6.4% p.a.), which clearly demonstrates how this performance is at the top end of potential outcomes. By comparison, the median return for all 10-year periods was cash +2.6% p.a.
The problem with averages or median statistics is that they assume we can live through many different such paths and that ultimately we get to experience the average outcome. But the reality is that each member only gets to experience one path over their lifetime. In an uncertain world, a fund’s investment strategy needs to factor in the range of potential outcomes when focusing on delivering the most optimal outcome for each member.
The broad industry direction towards low-cost, passive exposures with large allocations to equities and equity-like assets is strongly at odds with our own investment beliefs and approach. These are founded on the view that the best way to achieve a given objective is to minimise the range of likely outcomes around that objective. This means aiming to construct a portfolio with the same expected return, but with less volatility (the dispersion of outcomes around the average or expected return). Careful selection of active management and the use of risk factor diversification reduces overall portfolio volatility, and therefore improves the probability of achieving a given return objective (by reducing the range of possible outcomes).
Chart 2 shows the allocations we would generally recommend for a well-diversified portfolio, with a similar return objective to a 70:30 reference portfolio. The key difference being that our model portfolio holds 40% in equities, versus 70% for the reference portfolio and so as a result it has a meaningful exposure to a number of other risk factors.
Chart 3 overlays the performance of our model portfolio onto the return paths of the 70:30 reference portfolio. The range of outcomes for our model portfolio is much narrower, with cumulative excess returns over cash ranging from +6.2% p.a. to +0.2% p.a., compared to +6.4% p.a. to -1.5% p.a. for the reference portfolio, consistent with our beliefs and expectations. If we can narrow the range of potential paths, particularly limiting the negative ones, then we can significantly improve the chances of an optimal outcome.
Considering the last 10 years only, on a pre-tax, pre-fee basis, the reference portfolio delivered 10.1% p.a., while the model portfolio delivered 10.2% p.a. We estimate that, after superannuation tax and manager fees, this equates to 8.7% p.a. and 8.5% p.a. respectively. It is worth noting that both strategies have performed well during this exceptional period, with only a marginal net difference. This is not something that has been widely picked up on by the industry.
With the median outcome for a Balanced superannuation option of 8.5% p.a. over the same period (as per SuperRatings data), both the reference portfolio and the model portfolio have produced results close to the median. However, the reference portfolio’s volatility is around 8% p.a., while our model portfolio’s volatility was 6% p.a.
The higher fees needed to invest in the more diversified portfolio could explain the majority of the difference in outcomes over the last 10 years. However, if we only focus on the cost of active management and diversification, then we miss the true value of these portfolios. The superannuation industry has become increasingly concerned with peer-relative performance, and while there is always merit in judging oneself, the industry needs to ensure that any analysis of relative performance is multi-dimensional.
On a net basis, the more diversified portfolio has outperformed by approximately 50bps, with a volatility of 6% p.a. versus 8% p.a. for the reference portfolio, with significantly lower losses when the tech bubble burst in 2002 and during the GFC. Indeed the worst 12 month return for the reference portfolio was -27% during the GFC, whilst the model portfolio’s loss was 22%.
Indeed, there are also a range of qualitative factors that should be taken into account in considering the merits of a superannuation fund — an assessment of the quality of a fund’s governance, the level of resourcing, the clarity of investment beliefs, the articulation of objectives and the appropriateness of the investment strategy adopted.
Whilst it is not nearly as easy to assess these “softer” factors as it is to look at past performance and costs, ultimately these factors will have a much stronger bearing on a fund’s future outcomes and so they should have greater prominence in any assessment. One way to achieve this is to use a “quality scorecard”, which identifies a number of factors (both “hard” and “soft”). This is something we have developed for assessing any gaps in the approach of our clients.
In good times, returns rarely reflect the true risk of the underlying investments, and the last 10 years is a great example of this. While the GFC remains well within living memory, we are concerned that the remarkable outcomes delivered by a lower cost, passive approach is a pervading current influence on investment strategy, product strategy, and policy in the Australian investment industry.
Regulators and investors need to be able to judge outcomes not simply by considering relative returns and costs, but by reflecting on a broad range of considerations, just as we do when we seek to identify skill in an investment manager. This should involve a more balanced assessment of the factors that contribute to future investment performance – these are principally to do with the governance of the organisation, the quality of its people, its culture, and its approach to investing. Past returns and cost outcomes can then be considered more appropriately within this broader context. We’ll create a far better investment industry in Australia if we can get this balance right.
| Title | File Type | File Size |
|---|---|---|
| Let's get the balance right | .4 MB |
