During 2018 the Productivity Commission, at the bequest of the government, undertook a study of the efficiency and effectiveness of Australian superannuation funds.
Being very clever economists, they identified several grounds for improvement with perhaps their most insightful observation being that if those who invested in the worse performing funds had invested in the better performing funds, they would have a lot more money.
Such a brilliant observation does not deserve to go unnoticed, and so APRA picked up on it and developed a method to identify the underperforming superannuation funds. They have now been using this method for several years and during this time have forced the closure of many of the worst performing funds or their amalgamation with a better performing fund.
More recently, the Grattan Institute has recommended that APRA should extend its approach to account-based superannuation funds to get a similar performance improvement. Indeed, they went one step further and recommended that a list of ten preferred providers of account-based superannuation products should be drawn up and promoted to retirees as the best place to invest their superannuation balances.
The proposal presented by the Productivity Committee has great intuitive appeal. After all, members with a total of say $2 billion would be $40 million better off if they invested in a fund that returned 10% rather than one that returned 8%. Yes, but then no. Nobel Laurate Bill Sharpe pointed out to us well over 30 years ago that investing is a zero-sum game.
Another way of saying this is that for someone to outperform, someone else has to underperform. Hence, you cannot take funds from a losing fund and give it to a winning fund and assume that everything remains unchanged, including the return earned by the two funds.
In order to see this, let us return to our previous example but assume that the two funds mentioned were the whole market and each had $2 billion in funds. One returned 8% and the other 10% and so the market return was 9%. Now assume that APRA had got rid of the underperforming fund and that all $4 billion is now invested with the better performing fund, which now gets the market return of 9%. So, what have we achieved?
Well, we have certainly improved the returns of those that would have otherwise been invested in the lesser-performing fund, but this higher return has come at the cost of those that would have been invested in the better-performing fund. The fact is that previously we had winners and losers. But if we get rid of the losers from whom the winners were previously profiting, the previous winners are made worse off.
The bottom line is that the economists at the Productivity Commission showed a lack of understanding of how financial markets work when they assumed that you could create wealth simply by moving people from the lesser-performing funds to the better-performing funds. By moving funds from one manager to another you will change the dynamics in the market. The prices of the assets will be different, as will be the returns realised by the various funds. If the returns on the whole market remain unchanged (as was the case in our example), then there will be some winners and some losers but net no wealth has been created.
There are two other considerations that need investigation:
1. Can we expect an overall improvement in the performance of markets simply by closing down the recently poorer performing managers?
This is not entirely impossible as ‘better’ managers may result in capital being allocated to companies that use it more productively, resulting in faster economic growth and better market performance.
In our previous example, assume there was an improvement in allocative efficiency and so the market return was not 9% but was now 9.5%. In this case, the closure of the poorer performing fund would result in an increase in new wealth of $20 million. Those who would have otherwise invested in the poorer-performing manager would be $30 million better off and those already in the better performing managers would be $10 million worse off.
Unfortunately, the evidence does not support the argument that active managers contribute to improved allocative efficiency. Thus, it is likely that APRA’s attempts to rid the industry of inferior managers is only likely to result in wealth being moved between individuals without any additional wealth being created.
2. Can we identify the ‘poorer’ managers simply based on an analysis of past performance?
If we cannot do this, then there is absolutely no basis for excluding funds on the basis of past performance.
In order to address this question, we need to consider whether there is strong persistence in the performance of a fund over time. There have been numerous studies of such persistence through time over many markets. However, before we move to consider the empirical evidence, we should again take some guidance from our friend, Bill Sharpe.
Bill found that it took up to 40 years to identify whether a fund manager lacked the skill for the job. This is an extraordinary amount of time to identify whether an individual (or an organisation) has any special talents in the area in which she/he works. However, there is a good reason why this is the case: the skill-to-luck ratio that determines investment outcomes is extremely high and so it takes an inordinate amount of time/data when using historical evidence to separate skill from luck.
Although APRA makes its determinations based on eight years of data, there are clearly high risks associated with ‘terminating’ managers on this evidence. What is more, we will never know because in most cases the future performance of funds terminated is not measurable.
Now turning to the empirical evidence, we prepared a paper for ASIC about 20 years ago that surveyed the findings in a large number of papers on the persistence in investment performance. We found that there was little evidence to suggest that past performance was a good predictor of future performance even when measured over four-year periods.
Obviously the eight years used by APRA is an improvement but even then, expectations of its adequacy would be low. The results of persistence studies conducted over the last 20 years support our previous conclusion that APRA’s approach would likely be inadequate to identify funds with poor future performance, leading to the oft-stated phase that past behaviour is not a good predictor of future performance.
However, it does not stop there as additional indicators can be used to forecast future fund performance. Presumably some of these indicators are used by those that rank managers/funds for a living. So, how do they perform? Well, the evidence is not compelling and suggests that even the ‘experts’ are unable to consistently add value by their selection methods. Hence the seemingly rational suggestion by Grattan to have a ‘top ten’ list of fund managers is likely to result in more pain than gain.
Emeritus Professor Ron Bird (ANU) is a finance and economics academic and former fund manager.