Search for

Survival figures

31 Dec 08

If your investment vehicle is not working, you can merge it into one that is and its poor results simply vanish. Alan Dick looks at how investment houses can dress up the performance figures

by Alan Dick

How’s this for small print? “The information contained herein, not guaranteed by us, has been obtained from sources which have not in the past proved particularly reliable.”

You may have to refer a client to a financial adviser. However, much of the data used by advisers to support their recommendations is misleading due to little known statistical anomalies such as survivorship bias.

Investors and their advisors typically use performance data from commercial databases such as Morningstar, Lipper or Financial Express. Unfortunately, much of this data needs to be taken with a pinch of salt.

The data providers do their best with the information available, but the published data can be misleading.

It has been said that 90 per cent of investment funds can legitimately claim to be in the top quartile (25 per cent) of their peer group. How can this claim be valid? The answer is simply that the other 75 per cent of funds no longer exist.

This is an important concept when comparing investments, whether individual funds or peer group averages. When a fund closes, its entire performance history disappears from record. The result is that the reported returns only show the performance of the winners. Therefore, returns may be substantially over-stated. This is known as survivorship bias.

One study of the period 1962 to 1995 found that, of the 2071 funds in the study 725 closed.

How does this affect investors? Well, imagine an investment house, Old Planet, runs three UK equity funds (Pluto, Mars and Venus). Assume these funds represent the entire UK equity sector so their combined performance represents the sector average.

Three investors Tony, Gordon and Alistair each invest £1,000 in one of the funds.

Tony selects the Pluto fund, a medium risk fund that invests in a similar way to the FTSE All Share index. The manager’s bonus is based on his performance relative to the index so he is well diversified and never takes big bets on any individual stock.

Gordon and Alistair prefer a gamble and select the Mars and Venus funds which are both higher risk “alpha” or “special situations” funds where the manager is free to invest as he chooses. If the manager’s stock selection is good, the fund will beat its benchmark by a substantial margin, in which case he will earn an obscenely huge bonus. However, if his choices are poor, the fund will underperform its peer group by a large margin and he will unfortunately only earn a very large bonus.

Clearly there is significant incentive for the manager to take a big gamble with other people’s money in this hypothetical investment world.

Over 10 years, Tony achieves growth of 5 per cent a year from Pluto while Alistair’s Mars investment manages 10 per cent a year. Venus loses 5 per cent a year for Gordon during the first five years then closes because the poor performance is starting to make Old Planet’s figures look bad.

As an investment house, Old Planet aims to have most of its funds in the all-important top quartile as this gives the boys and girls in marketing something to shout about. Venus is merged with Mars at the beginning of year six. Venus no longer exists and Gordon’s investment has been moved to Mars.

You can see from the table that after five years, Pluto has returned 5 per cent a year while Mars and Venus have managed 10 per cent a year and -5 per cent a year respectively but what of the situation at the end of year 10? As a result of the merger of the Venus and Mars funds the database shows Gordon, who originally bought the Venus fund achieved growth of 10 per cent a year (remember that Venus no longer exists and its performance history has been replaced by Mars) when in actual fact he has really received a return of only 2.2 per cent a year over the 10 years.

It is also worth pointing out that the sector average of the three funds over the 10 years in question was 6.2 per cent a year. However, the merger of Venus and Mars has miraculously increased the sector average return quoted by the databases to 8.5 per cent a year.

Surely this type of blatant performance laundering doesn’t really go on in the real world? At the end of the 1990s, a few weeks before the Internet bubble burst, Gartmore launched an internet fund called Tech Tornado. Investors lost around 70 per cent of their money almost overnight.

A quick look at any of the investment databases will reveal that this fund no longer exists. It was merged with the Gartmore UK & Irish Smaller Companies fund. Investors looking at Financial Express for performance data on their fund will find that they have only lost 3.23 per cent a year over the last five years and their fund has managed growth of 10.22 per cent a year since launch, almost double the return of the average UK Smaller Companies fund.

They are probably a little confused as to how this can be true as their fund is still worth a fraction of the amount originally invested.

A more recent example of this laundering strategy is New Star. At the beginning of 2008, New Star’s UK Equity funds were languishing near the bottom of the performance tables. According to Investment Week, the UK All Companies sector returned 2.08 per cent in 2007 but the New Star Special Situations fund managed to lose 15.4 per cent. This poor performance had continued for some time so New Star merged the fund into its UK Alpha fund at the end of February 2008. All mention of the UK Special Situations fund has now been erased from the records and the woeful performance has been replaced by the better UK Alpha fund which has beaten the sector average by around 2 per cent a year since launch in November 2001.

In November 2008, New Star announced that Stephen Whitaker was to be replaced as the manager of the UK Growth fund due to a sustained period of poor performance. The new manager may, or may not, be able to turn the fund around so it is not beyond the realms of possibility that this fund will also be merged into another fund.

I am not picking on Gartmore or New Star as all fund managers use this trick.

Leading academic studies estimate survivorship bias, and other statistical anomalies, contribute to the returns from equity funds being overstated by more than 2 per cent a year. Thanks to compounding, it can overstate the cumulative returns over 15 years by about a third and over 20 years, by 45 per cent.

If you are referring a client to an investment adviser, make sure that they are aware of these issues and can explain the impact they have on their recommendations. Unfortunately, most advisers are completely unaware of the problem, in which case, find another adviser. n

ALAN DICK is a certified financial planner and partner with Forty Two Wealth Management. He is one of the founding members of EBIS (Evidence Based Investment Solutions) a group of independent investment professionals.

Have your say

Page No: 28


Related Articles

Practice Web Tower (link opens in new window)Advertisement