Recently in New Zealand a few star managers have had brilliant runs, easily outperforming markets since inception.

Why on earth would an investor not want to use these funds for their New Zealand share allocation?

We have a lot of respect for what star performers have done.  It’s not easy to produce exceptional performance.  But for all their glory, the websites of the star performers provide us with the answer to the question above… “Past performance is no guarantee of future returns.”

When asked a question about a star investment manager we typically respond with a question of our own, “Would you believe that this is a very well studied issue?”

Many academic, peer reviewed papers have been written on the question of whether or not we can simply allocate money to a fund that has a good record, and expect to beat the market in the future.  It’s a topic that has been researched across different time periods, different countries and different asset classes.

The findings of these studies are just as the disclosure statement suggests – past performance is no guarantee of future returns.

Perhaps we would put it more strongly: the evidence shows that past outperformance tells you next to nothing about future outperformance.

Below, we summarise two of the many papers on this subject, one from the United States and one from New Zealand.

In 1997, Dr Mark Carhart published “On Persistence in Mutual Fund Performance” in the Journal of Finance.  This paper addresses the exact question posed above.  Does the good performance of investment managers persist?  Can I select a manager with good records and expect to beat the market?

To answer this question, Carhart uses a comprehensive database of 1,892 equity (share) funds from the period 1962 – 1993.  Regarding the comprehensive nature of his data, he states, “the data… include all known equity funds over this period.”

The author summarises his conclusions on the persistence of investment managers’ performance this way:

“Common factors in share returns and investment expenses almost completely explain persistence in equity mutual funds’ mean and risk-adjusted returns… The only significant persistence not explained is concentrated in strong underperformance by the worst-return mutual funds.  The results do not support the existence of skilled or informed mutual fund portfolio managers.”

In other words Dr Carhart is saying that, once he accounts for types of investment risk taken by the manager, investment managers beat the market in the future only by chance.

In New Zealand there is only one published study we are aware of that analyses the persistence of New Zealand investment managers.  In 2006 Bauer, Otten and Tourani Rad published “New Zealand mutual funds: measuring performance and persistence in performance” in Accounting and Finance.

This study reviewed a sample of 143 mutual funds for the time period 1990 – 2003.  Their results show that performance in New Zealand mutual funds does not persist more than luck alone can explain.  The authors conclude:

“It has to be noted that the documented persistence in performance is mainly driven by icy hands, instead of hot hands indicating that funds that underperform (significantly negative alpha) in one period are most likely to underperform in the next period.  Investors should therefore avoid these funds.  However, evidence of persistently out-performing funds (significantly positive alpha) is absent.”

In summary, the paper finds no evidence that an investor can reliably outperform markets in the current period by selecting a fund that outperformed markets in the last period.

It is probably helpful to get away from the academic articles for a moment and just look at some practical examples.

There are businesses like Morningstar dedicated to rating funds.  Their proprietary research goes beyond mere performance and looks to identify quality in investment management and investment approaches.

Morningstar gives their best rated funds five stars, while the worst funds get one star.  If anyone is going to find a winning fund manager, it’s Morningstar.

Yet their record is in performance prediction is abysmal.  A recent study looked at the 248 managed funds that Morningstar rated five stars in 1999 to see how they had performed over the ensuing decade.

The chart below shows the story.  A decade later only 4 out of 248 funds were still five stars, and 87 funds had gone out of business.  The average return of a Morningstar five star fund over the 10 years was worse than the average return of all other funds put together.

The data here corresponds to a study by Paul Gerrans published in the academic journal Accounting and Finance entitled “Morningstar ratings and future performance”.

Dr Gerrans found that:

“By far the strongest case to mount is that of no significant difference between relative rating and future performance…. if an investor is looking to the star for guidance, this analysis suggests that they might be better served examining information on fees, which over the period of analysis appear greater on average than ratings differentials and are far more certain.”

So we can see Morningstar is likely no guide.

Below we look at all the funds monitored by Morningstar from 2004 to 2008 in the US Large Cap Blend asset class and ranked them from the very best performing fund to the very worst performing fund.  Thanks to Index Fund Advisors (www.ifa.com), we can see how this looks in graphic form using the top chart from the figure below.

The bottom chart shows the performance of those same funds from 2009 – 2013, keeping the same ordinal ranking as the top chart.  There’s no real pattern, other than perhaps, on average, the badly performing funds in the first period did better in the second period.

It’s no wonder top academic researchers have told us that past performance does not persist and is no guide in selecting fund managers.

In another study, Vanguard ran a test to show the practical implications of this research, where they compared two strategies that investors may employ.  The first they called “performance-chasing” and the second, “buy-and-hold”.  For the performance-chasing strategy they randomly selected a fund in the top 50% of peers over the past three years. If the fund ever fell below the top 50% they would replace it randomly with another fund of the same asset class in the top 50%.  In other words, this strategy was always holding a fund in the top 50% of peers based on three years of historical data.   For the buy-and-hold strategy they picked a fund at random, regardless of performance history, and held it through thick and thin.

They ran this experiment over and over again, across several asset classes.  Below we see the median result for all nine asset classes they tested.  Buy-and-hold crushed performance-chasing.

The Vanguard study picked funds at random.  In the real world, funds aren’t picked at random.  Pension funds, for example, spend quite large amounts of money on consultants to help them pick the funds that are likely to be winners in the near future.  Another study published in the Journal of Finance looked at pension funds in the United States.  The study showed very clearly that the investment managers selected by these pension funds had very impressive track records.  They produced excess returns (fancy talk for beating the pension’s existing funds) by 4% – 9% in the years before the pension funds selected them.

The study was on how the funds performed after they were selected.

The answer?  The fired funds beat the hired (or selected) funds!  The first year the fired beat the hired by 0.49%; the second year by 0.88% and the third year by 1.03%.  Why?  Because past performance is no guarantee of future performance.  The results of the study are illustrated below.

What can we learn from this evidence?  If persistence in performance is mostly random, then picking managers on the basis of good track records is likely to:

• Increase costs and therefore reduce our expected return
• Reduce the certainty we achieve that return

Neither of these results are in the best interests of investors.

None of this is to take away from the achievements of star managers with great recent performance.  But it does mean that we don’t believe that their performance, impressive as it is, forms a basis on which we can conclude it is likely to persist in the future.

So what qualities and characteristics can you use to prudently select investment managers?  We’ll cover that next time.