Hello, everyone. Welcome to Retirement Puzzle, brought to you by the Monash Centre for Financial Studies. I’m Ummul Ruthbah, one of the hosts of the podcast where we explore the current and critical trends in the pension sector and try to explain how they impact us all.
In this episode we are going to talk about the self-managed super funds, SMSFs for short, who are an important part of the Australian superannuation system. As of last September, they had about $A728 billion assets under management which was about 25 per cent of our total super assets.
Despite their huge size, not much information is available about their performance, persistence, preferences or prejudices. Researchers at MCFS have been working with a large set of self-managed super funds portfolios’ day-to-day trading data covering eight years which puts them in a unique position to fill in some of this knowledge gap.
Today I have here with me two of my esteemed colleagues – Dr Nga Pham and Dr Bei Cui from MCFS who will share their first-hand knowledge of the trading behaviour of the SMSF investor community. Hello Nga.
Nga Pham: Hi Ummul.
Ummul: Hi Bei.
Bei Cui: Hi Ummul, thanks for inviting us.
Both Nga and Bei have Phds in Finance. Nga is also a CFA charter holder. Their areas of research include sustainable investments, corporate governance, superannuation and retirement.
Ummul: Nga, let me start with you. MCFS just finished a research on the performance of self-managed super funds. Could you please give our listeners some ideas about these funds that you studied?
Nga: Hi Ummul. Thanks for the question. I have to say that the research on SMSFs with SelfWealth Ltd, our industry partner on this project, was the highlight among some of our recent projects. We have written two research briefs and a white paper on the performance and investment strategies of SMSF investors. They are all on our website.
The SelfWealth database allows us to examine the daily trading behaviour of more than 60,000 SMSF investors with their Australian equities portfolios. Our analysis started with over 27,000 portfolios with a total value of $12 billion in 2012. Our sample expanded to more than 43,000 portfolios as of mid 2020, right after the March COVID panic of the market with the total value of over $25 billion.
Ummul: What was the average size of the portfolios and how did they evolve over time?
Nga: On average, the value of the portfolios increased from around $470,000 in 2012 to about $730,000 in 2020. The median value was just half of the mean values. This means that the average values of the portfolios were driven by some very large SMSFs in the sample, the multi-million dollars of assets. Yes, I mean some portfolios were quite large. The largest one in the database held almost $390 million of Australian financial instruments in 2020.
Ummul: What are the different assets these self managed super funds invest in?
Bei: SMSFs can invest in a wide range of assets such as stocks, including both in domestic and international markets, bonds, options and some types of real assets. However, our study only examined the performance of SMSF investors of their domestic financial assets portfolio including Australian equities, listed investment companies, ETFs, options, warrants some fixed income assets traded on the ASX. However, equity was the biggest components, which constituted 79 per cent to 89 per cent of the non-cash financial asset portfolios. So we only looked at a part of an SMSF portfolio, and I think that is the most exciting part.
Ummul: Could you please share with us the overall performance of the SMSFs in your studies?
Nga: It was not a very rosy picture that we have seen. Overall, when we analysed this whole SMSF investor community in the database, we found that on average SMSF investors did not beat the market. We created an index to represent the performance of the SMSFs in the sample. To put it in dollar terms, $1 invested in the S&P/ASX200 from mid-2012 would give you almost $2.50 up until the peak before the covid-plunge in Mar 2020. The performance of SMSFs was worse than that. $1 invested in 2012 would become only less than $2.20 just before the pandemic.
SMSFs also had lower volatility than the broad market index during the years we examined. However, even when we accounted for the lower risk, the risk-adjusted returns of SMSF index was still lower than the market.
Basically, it just means that for most of these SMFSs, they would have been better off investing in an index fund that gives them broad exposure to the Australian market than managing their own portfolios.
We also compared the performance of the SMSFs index with a number of broad market style-based indices. If you are interested in this comparison, we have the details in our research briefs on the website.
Ummul: Did you find any relationship between size and performance? The productivity commission found that the large portfolios more or less kept up with the APRA regulated Funds. It is the small ones that performed poorly compared to the large SMSFs. Did you find any relationship between size and performance like this in your sample?
Nga: Yes, we divided portfolios with at least 20 stocks into four groups based on size. The largest group had assets of more than $1.65 million. The smallest group had less than $400k. We find that large funds performed better than small funds during the first period of our analysis. In the subsequent period that includes the pandemic time, the performance of the groups were relatively similar. However, the performance persistence of large funds over the whole study period was better than small funds.
Ummul: While reading the report, what I found very intriguing was that even though on average the SMSFs in the sample did not perform very well, there were some among them who did quite well. Bei, Could you please tell us how you found those patterns?
Bei: Most SMSFs in our study indeed failed to keep up with the market however we have seen persistence among the best SMSF performers in the community. It suggests that there was some level of wisdom within the crowd among the top performers and in that spirit we developed an approach to use historical performance indicators to identify past outperformers. Our expectation was that these portfolios would continue to do well in the future.
We have backtested various risk and performance indicators to pick up the top performers each year and indeed we saw that these top performers still performed well in the following year, which validated our conjecture.
Ummul: It is an interesting contradiction since there is a belief that past return is not an indicator of future returns.
Bei: Yes Ummul. That is an interesting and important twist. The top performer SMSFs in our study exhibited a higher level of persistence in their performance, compared to the crowd.
Ummul: Can you explain to us in detail how you selected these top performers, which are labeled as “winners” from you sample Self Managed Super Funds community and how well did they perform compared to others?
Bei: Overall, the resulting strategy is based on two underlying principles- the first is that a reliable winner portfolio must be in the top decile compared to its peers and the market in all risk-adjusted indicators and the second one is that it must not be in the top decile of risk indicators compared to its peers. We repeated the selection of winners every financial year based on past years’ performance.
We then tracked the performance of the selected winner portfolios over time. The resulting SMSF winners yielded 9.3% annualized return compared to 6.4% of the S&P/ASX 200 total return index. At the same time, the SMSFs winners we selected had lower standard deviation compared to ASX 200.
Ummul: That is impressive! Did they have a more diversified portfolio compared to the Self Managed Super Funds community?
Bei: The number of different stocks held collectively by winner portfolios ranged from around 400 up to 800 per day during the eight-year timeframe of our study. ASX 200 and ASX 300 stocks comprised between 73 per cent and 79 per cent of winners’ holdings respectively. Although individual SMSF winners typically held a variety of stocks, their holdings collectively were highly concentrated on certain stocks. The average dollar amount invested in the 20 most popular stocks held by SMSF winners accounted for 53 per cent of the dollar amount of total holdings, over the studied period.
Ummul: Did you find any preference for particular sectors among your winners??
Bei: Yes, on average, compared to the weight structure of S&P/ASX 200, SMSF winner portfolios’ top 20 holdings were overweighed in Healthcare, Industrials, and Information Technology and underweighted mainly in the Financials sector.
Ummul: What were some of the popular stocks among your winners?
Bei: The top three stocks they held across the sample period were CSL limited, Commonwealth bank and Westpac. For CSL limited, the daily average weight in SMSF winner portfolios was 8.85%, which is 5.47% higher than the weight in ASX200 index.
Ummul: Did you find any difference in the investment strategy of those you call winners and those of the losers or worst performers?
Bei: We have seen that there is a lot of overlap between what the best and worst performers held over the whole sample period. However there were differences in their timing and the weights of stocks in the portfolios. We believe that winners outperform the market because they are overweighted the right stock at the right time while the worst performers underperform because they are overweight in the wrong stocks. So when we developed an investment portfolio that selected the top overweighted stocks in the winners’ portfolios and filtered out stocks most overweighted by the worst performers, the portfolio significantly outperformed the market.
Ummul: Did you consider the transaction costs of rebalancing the portfolio?
Bei: Yes, we included transaction costs in the analysis and after including transaction costs. The risk-adjusted return performance of the winner portfolio stayed strong after we included the transaction costs. Only if the cost increased to an unrealistic level of 250 to 270 basis points would we see the outperformance eroded. This suggests that the strategy produced superior performance compared to the market, net of transaction costs.
Ummul: So there is indeed ‘’wisdom’’ within the crowd of SMSF investors.
Now I would like to know what happened during the Covid -19 period. How did the Self Managed Super Funds community behave and perform? Did you observe any panic selling during the market’s trough in March 2020? Any observation on how SMSFs changed their investment priorities and asset allocation strategies during the first phase of Covid-19?
Nga: The analysis of the investment strategies of these self-managed super funds during the onset of the COVID pandemic was very interesting. We have seen no evidence of panic behaviour. What we have seen was a swift and huge decline in value of many portfolios but then a very fast recovery for most of them.
Ummul: What was the extent of these decline in value and how much of it was recovered?
Nga: Within the first three weeks of March, on average, SMSF portfolios reduced by almost 30%. The total decline in dollar terms of the whole community was more than $7b. There were SMSFs that lost millions of dollars. The biggest drop of value we saw was more than $100m. Who can sit on this money and watch it go by millions of dollars every day. It was a very tough time.
However, within three weeks after that, until mid-April 2020, most funds had recovered around 90% of their pre-March value.
The reason that these SMSFs could come out of the market panic drop without hurting themselves much was because there was no panic selling as seen in the broader market.
Ummul: So there was no panic sale. Did they buy during those weeks of panic?
Nga: There was a net buy in the community of almost $50m in January 2020. And the value of net buy of the community doubled in Feb and tripled in Mar to more than $140m. Perhaps buying the dips has helped these SMSFs to quickly recover from their loss. They were in a good position when the market bounced back after March 2020.
Ummul: These are some very fascinating findings. Thanks, Nga and Bei for your time and for sharing your research insights of this less explored investor community.
We learnt that there are smart investors within the SMSF community. More importantly, the results reveal that to optimise a crowd-sourced investment strategy, investors should examine the strategies of both the best and worst-performing portfolios.
One significant potential real-world application of this research, particularly for share trading platforms such as SelfWealth and those run by the big Australian banks – all of which have ready access to large investor databases, is that they could use the findings, for example, to help guide the refinement and development of exchange-traded share funds and other crowd-sourced investment products.
Thank you for listening to The Retirement Puzzle from Monash Centre for Financial Studies. If you’ve enjoyed it, please tell your colleagues and friends about us. You can subscribe to our show anywhere you listen to podcasts. And don’t forget to leave us a review. If you want to collaborate with us on retirement related issues, please get in touch at email@example.com.