How QSuper Halved Risk Without Losing Return

Inspired by risk parity strategies and illiquid-focused endowments, Australia’s second largest superannuation fund is pioneering a radically different approach.

Damian Lillicrap
Head of Investment Strategy, QSuper

In the years following the financial crisis, the QSuper investment team executed a dramatic shift in portfolio exposures, slashing equities while more than doubling fixed income and infrastructure. The most controversial aspect was a new reliance on long duration bonds: “the industry seemed to think we were mad,” recalls strategy head Damian Lillicrap.

Yet the subsequent results have been hard to ignore. This new, equity-light portfolio has not only kept up with its equity-heavy peers through an extended bull market: it has outpaced them, and done so with half of the volatility.

In today’s climate, as investors become increasingly concerned about corrections, will others follow in QSuper’s footsteps? “We do see other investors reducing equities and increasing bonds more as time has gone on, albeit not to the degree that we moved,” notes Lillicrap. “The bull market since the GFC has created complacency, which is not good when you consider the growing risks out there.”

Q: Damian, thank you for joining us to talk about some of the innovative work that QSuper has been doing. Before looking at the latest developments, take us back to the roots of this ‘risk balanced’ strategy. How did it come about?

To understand the change to our strategy, it’s important to think about the conditions when the QSuper investment team was established in 2009, at the tail end of the GFC. The portfolio had just experienced a loss of about 25% in value and the market was still highly volatile. On a relative basis, these returns were ‘good’, and other funds had lost 28% on average; on an absolute basis they were obviously awful for the members. When the team was set up we were reflecting on these losses in quite a fundamental way and asking: can we do something different – something better?

QSuper’s investments had previously been outsourced to QIC [Queensland Investment Corporation, where Lillicrap had been head of portfolio design]. The move from QIC to QSuper, in a nearby building, was less than 100 metres. But in other ways the move was huge. Before the move, I believed I had one client [QSuper] that I felt I was doing a great job for; afterwards I realised that we had five hundred and fifty thousand clients [the members] that we could do a far better job for. The sense of responsibility was much higher.

At the time, we had what might be considered a typical 70/30 style portfolio: more accurately 60-65% in listed equities, plus equity-type exposure in private equity and infrastructure. From a risk perspective this translates into about 90% of portfolio risk from equities. The typical drawdown for that type of portfolio in the GFC was about 28%. We did somewhat better (-25%), because we had introduced dynamic asset allocation after the tech wreck – it was quite novel in the Australian super landscape at the time. This really showed us the limits of DAA: it had improved the outcome, we had made the calls about as well as we could have hoped, but it hadn’t solved the fundamental problem. It wasn’t the answer.

Q: So your team was on the hunt for a different way of doing things – something that would perhaps avoid these very large drawdown-type events. What was the next stage?

We stepped back, started with a blank sheet of paper and considered what investors were doing globally. The approach we settled on was quite heavily influenced by two groups: university endowments and risk parity managers. We had been developing relationships with endowments for some time and had already started a journey towards unlisted assets. We had also engaged with some risk parity managers but only as a potential investment; this was the first time we had considered incorporating aspects of their thinking at whole of portfolio level.

After studying the endowments we committed to increasing unlisted asset weights. Meanwhile the risk parity approach led us to want to diversify away from a dominant risk factor, equities, to develop a portfolio that was more robust to a range of economic outcomes.

In practice, this meant that we roughly halved equities from 65% to 35%, increased fixed income from 10% (in a relatively standard index) to 25% (with a much longer duration), took infrastructure from 5% to 17%, kept property around 8% (the opportunities in infrastructure were more attractive than in real estate) and built private equity to about 5%. In the last couple of years we’ve also introduced trend-following, which is now at 5% of the portfolio. We still have a dynamic asset allocation strategy.

Q: How difficult was it to persuade all the relevant leaders and stakeholders to take this approach, especially the changes on the fixed income side?

We wanted exposures to the type of fixed interest that gives better diversification to portfolios: high quality government bonds. The industry seemed to think we were mad. There was a widespread consensus out there that yields would rise. Yet when we looked around the world at debt levels and growth levels we thought differently: we weren’t betting that yields would fall further, but we estimated that this may be a ‘new normal’. Most people are not very familiar with high duration bonds, compared to conventional fixed interest and equities. Their roll-down and curve effects are not always appreciated. When the investment committee and the board considered the improved member experience, and the analysis around the bond outcomes, they supported the analysis.

Q: So how has the fund performed? Have you fallen behind in what has largely been an extended equity bull market?

Over the ten-year period between 2009 and 2019 we were the highest performing super in the country (+8.6%pa), with Balanced strategies being the standard Australian default. Yes, equities have had a bull market (11%), but our bonds have also delivered (8%), and infrastructure and private equity have both performed exceptionally well (~15%pa). The aim wasn’t to get the highest returns, but what is clear is that reducing risk has not cost us returns. Since we changed the strategy, in 2011, we have had half the volatility of the typical Australian super fund. While volatility is not a perfect measure of risk, we do think there is material information in relative volatilities, indeed more information than there is in comparing long term returns as a proxy for future relative returns.

Q: What changes have you made recently, or are you thinking about making, given the current market conditions and opportunities?

We still have quite a dynamic approach to asset allocation, even though we recognise that it only changes outcomes by a relatively small margin. This leads us to selling asset classes as they get more expensive and buying as they get cheaper, which has generated quite a bit of activity given recent volatility.

With our longer duration bonds we seek high quality government exposure that has the capacity to rally if there is a growth shock. As yields around the world have compressed to zero and turned negative there are a dwindling number of countries that we are prepared to invest in. The fixed income portfolio is therefore dominated by Australian and US bonds, that still have the capacity to offer material downside protection.

In illiquid investments, we absolutely recognise that we’re not going to get the same returns in the future that we’ve had in the past – our results there have been top quartile, and through a very strong period when everything has benefited from yield compression. But we are still quite confident that the returns here will meet our internal objectives. One reason that we’ve been getting those top quartile results is that we’ve been happy to step away from standard approaches. For example, we were early to adopt things like very specific mandates with right of refusal on individual assets being purchased: filtering out assets we didn’t like the look of has been a big contributor to performance. Today, our focus is on “non-core” markets: we’re looking beyond the core five or six markets that have dominated the portfolio and are looking at emerging markets as well as non-emerging markets that we haven’t used. We’ve been travelling to Latin America, India and China, trying to get a deeper understanding of the countries that we know will dominate indices in ten or twenty years’ time.

Q: Do you see other pension funds moving more in your direction? If not, why not?

We do see other investors reducing equities and increasing bonds more as time has gone on, albeit not to the degree that we moved. To be fair, it would have been hard to follow us - bond yields have fallen, making it more challenging to swap large amounts of equity risk into bonds.

However, I still don’t think there’s been as big a focus on risk among pension funds globally as there ought to be. We have delivered great returns with about half the risk of the traditional approach since we changed strategy eight years ago. Theory would suggest that as our approach has twice the Sharpe ratio it should be twice as desirable as the traditional way. However, the bull market since the GFC has created complacency, which is not good when you consider the growing risks out there. In the event of another GFC, QSuper’s portfolio should incur less than half the losses that we had last time, whereas most funds haven’t changed things that much. If anything is going to have members asking for change it will be repeated large losses. We hope we are ahead of that cycle.

As investors take further steps to improve their diversification and resilience late in the cycle, we continue to observe rising allocations to areas such as infrastructure and trend-following strategies. With QSuper in mind, we look forward to seeing whether risk parity-inspired approaches gain traction.