a leaked document obtained by investment week highlights the difficulty of using past fund statistics ' particularly over a three-year period
Three-year past performance figures are less reliable compared to other timeframes, an IMA report on past performance shows.
According to the Charles River Associates report, commissioned by the IMA, persistency of fund performance does exist, both among bad funds and good.
The report concludes that as persistency is not random, it is a useful tool in picking funds. Using performance information from Standard & Poor's augmented with other sources, Charles River found roughly a third of funds in the top quartile of any of four UK sectors during the past 20 years, when looking at 12-month past performance, could reasonably be expected to still be in the top quartile 12 months on.
If returns were random, the proportion of funds retaining their top-quartile performance in the UK All Companies, UK smaller companies, UK equity income and UK equity & bond sectors should be just 25%.
Importantly, the group's calculations, done on a net and gross basis from 1981 to 2001, take into account funds that have closed down, something previous studies have not always included.
The study included 508 funds available at the end of the period as well as 434 funds that were closed during that time.
Tracker funds, which were first marketed in 1988, were not included.
Of course, a third is definitely better than a quarter, but it is hardly a compelling proportion of funds. Quite simply in just 12 months, the wheels fall off a lot of funds, or at least the momentum that took them to the top of its peer group has eased.
Shown opposite is some of the plethora of supporting data published in the document. What it shows is the strength as an indicator of past performance of varying periods for funds listed in the all-encompassing UK All Companies sector.
It shows that when a fund which is top quartile over the past 12 months is bought, it has, historically, calculated using gross returns, had a 36.9% chance of still being there in 12 months time.
In the UK smaller companies sector, the corresponding figure is 39%, while in UK equity income it is 32% and in UK equity and bond sector 35%.
Using five, six and seven-year top quartile past performance as an indicator returns similar results. Using two and three-year top quartile past performance for funds in the UK All Companies sector as an indicator produces far less compelling results.
A similar trend can be identified in the UK smaller companies sector and even, to a far lesser extent, in the UK equity income sector. This dip in the effectiveness of the three-year track record of a fund as an indicator of future performance is surprising.
The report was never designed to look behind past performance to account for persistency, merely to prove it existed and has a useful part to play in stock selection. Nevertheless, theories accounting for this compelling statistical dip include the style bias of managers and funds coming in and out of favour due to market climates.
This dip in the persistence of past performance over three years has implications for advisers using three-year figures to select funds.
Many advisers have moved away from simple cumulative performance in their analysis of funds anyway.
When Investment Week contacted Bates Investment Services, Chase de Vere and Hargreaves Lansdown to discuss the results of the report each stressed the importance of looking at a range of different periods, both discrete and cumulative.
Anna Bowes, savings and investments manager at intermediary group Chase de Vere, also pointed out past performance alone can be deceptive. She gives the example of funds marketed using track records earned by a previous manager who has since left the group. Bowes said: 'We have always believed past performance is a useful indicator of future performance, but investors must be careful how they use it. You cannot just look at past performance figures without knowing what is behind them.'
Bowes said investors should understand the style of their fund, whether it is a value or growth-based fund, or a blend fund.
When the stock market favours value funds, as it does now, growth funds will struggle. This can explain why a previously good fund has slipped down the performance charts.
She said: 'Neil Woodford has always been a quality fund manager but there have been times when he has underperformed because he has stuck to his guns. When fund managers were pouring into technology stocks he did not, but since then he has done very well.'
Ben Yearsley of Hargreaves Lansdown, said the higher identifiable persistency levels in the UK equity income sector appear to confirm the advisers' belief that the sector boasts one of the highest proportion of experienced and long-serving fund managers. With names such as Woodford and Graham Kitchen at Invesco Perpetual, Bill Mott at Credit Suisse, Adrian Frost at Artemis, Jeremy Lang at Liontrust, Tony Nutt at Jupiter, and Toby Thompson at New Star, the sector appears highly competitive.
The report sought, under its mandate, to prove the usefulness of past performance in fund picking as well as the existence of past performance persistence. This was important to the IMA when setting Charles River Associates its brief, as it sought to prove the FSA's stance on past performance, that it is no guide to the future, is wrong.
The IMA hopes the regulator will revisit its decision not to include past performance in its online comparative product tables, something the FSA is not inclined to do.
The FSA refused to comment on the report by Charles River Associates, saying it has not yet received a copy.
The Charles River study showed greater gains can be made from persistency after accounting for initial and annual charges as this is likely to be reflecting that poor performing funds die so consumers are likely to incur further charges when they reinvest in another portfolio.
According to the study, a fund in the top quartile of the smaller companies sector, based on the previous 12 month investment period and holding it for a year, gives an average cumulative return over the 21 years (1981-2001) of roughly 20%.
This figure falls to 15% if one based the decision on a fund in the bottom quartile based on the previous 12 months performance.
By choosing a fund in the top, rather than the bottom, quartile, based on its previous 60-month performance and holding it for the following 60 months, produces a higher cumulative return of at least 4% in three out of four sectors, the report stated. UK smaller companies produces a return about 12% higher, UK Equity Income about 16%, UK Equity & Bond shows no difference in the return while the UK All Companies is around 4% higher.
Key document findings
• Persistence of performance is a widely observable phenomenon that can be economically exploited by retail investors.
• Taking charges into account does not negate persistence ' it actually increases the effect.
• Persistence is more than just a short-term phenomenon.
• Even short-term persistence can be economic where switching costs are low.
• Persistence is evident in the 1990s alone as well as the period as a whole.
• Results are robust within and across sectors
• Results are statistically significant.
• Three-year past performance worst indicator of future performance.
Source: Charles River Associates
To promote 'long-term investment'
Switching 'hard and expensive'
Smaller funds still packing a punch
To drive progress