Mary Holm: Don't glance in the rear-view mirror

It's Saturday night, and you've got a choice of two parties to go to. Do you pick the house where they had a great time last night, or the one where they had a quiet Friday evening? The choice isn't obvious. If the people in the first house turned on a good party once, why not twice? Then again, they might be all partied out. The folk at the other house, who rested up the previous evening, might be in better form tonight.

The situation is similar in investment. You might do well if you put money into last year's best performer. But quite often you won't. This is not news. Ads and literature about investments frequently warn that "past performance is not necessarily indicative of future results", or words to that effect. But many people take about as much notice as smokers do of warnings on cigarette packets.

Why? Because in other areas - sport, the arts, academic performance - whoever did well last time is quite likely to do well next time. But much research shows that the same does not apply to investment.

Before we go further, I'm not talking here about graphs of returns on different types of assets, such as shares, property and bonds, over a decade or more. The story they usually tell - that shares are the most volatile but also tend to produce the highest long-term returns, followed by property - is a valid story. What I am talking about is short-term comparisons between asset types, or between different markets or managed funds.

Whenever somebody says, "Property gave the biggest returns last year. Let's move your savings into property," or "This fund was the best in the market last year. Why not put all your new savings into it?", be wary, for two reasons.

1. Such comparisons are often trumped up. A company pushing a product is likely to choose a period that shows it in a good light. They may also compare a fund's performance with a market index that covers a different market sector than the fund, such as large companies only. Or the fund's investments may be riskier than the index. In that case, you would expect the fund to outperform the index in some years but not always. You need to appreciate the fund's riskiness.

Comparisons can also be misleading if one investment excludes fees and taxes, or includes reinvested returns, and another doesn't.

2. Even if the asset, market or fund did win last year, often the winner is no more likely to keep doing well than other alternatives. In some situations, it is more likely to do poorly.

Some examples:

* Comparing sharemarket performance in nine developed countries since 1974, we find that on six occasions the best country one year was worst the next year, or vice versa.

* The best performing industry worldwide in 2002, consumer staples, was the worst performing in 2003. And the worst in 2002, information technology, was the best in 2003.

* The table in this chapter shows that on four occasions in the last 10 years, the worst asset did best the following year, or vice versa.

Okay, I'll admit these are selected cases. There are also plenty of times when good or bad performance persists for a while. But there's no way of knowing, in advance, when this will happen.

Recent Australian research looks at the issue more broadly. Academics assessed about 100 studies done in various countries over 20 years. About half the studies found no relationship between good past and future performance. In some other studies, there was some relationship, but usually only in the short term. If you invested on the strength of that knowledge, you would be moving your money around frequently. That would be not only time-consuming, but any gains you made from the strategy would probably be eaten up in brokerage, fees and possibly tax on capital gains.

Overall, the Australians' conclusion was that past performance is not a useful guide for the future. And similar British research came to the same conclusion.

So, if you shouldn't switch to last year's best performer, how about being a contrarian, and switching to last year's worst? Sometimes that works well. But, according to the Australians, some research suggests poorly performing funds are slightly more likely to keep doing badly over the short term. Presumably that's because they include not only funds investing in areas that happen to be down, but also funds that are poorly run. And it's not easy for the layman to pick which are which.

Assuming you've chosen your long-term investments wisely, your best bet is to stick with them - although you should, of course, always keep them under review. That strategy is easier and cheaper than any other. And you've probably got just as good a chance of doing well in the future as you would have if you switched.

Why don't winners stay winners?
Fund managers that perform well in one period don't necessarily do well in the next period. Various reasons have been offered. Among them:

Risk
High performers are likely to be funds that invest in riskier shares or other assets and have had a lucky year. Because they take more risk, it's also likely that they will do particularly badly in an unlucky year.

Management style
Many fund managers follow a particular strategy that works well in some business environments, but badly in others. Some managers favour smaller companies, or high-tech stocks, or New Zealand rather than international shares. Some concentrate on growth stocks, which are recent high performers, while others favour value stocks, which seem to be cheap. In some markets growth stocks are king; in others, value stocks are king.

Luck
To the extent there is luck in investment choices - and there's probably more luck than many professionals acknowledge - no fund will always do well. It will be lucky some years; unlucky others. Over the longer haul, its performance might be about average. This is called reversion to the mean, and it happens.

Manager movement
To the extent there is skill in investment choices, the fund with the best manager will probably keep doing well. The trouble is that highly successful managers tend to be poached away by other funds.

In New Zealand in particular, where fund manager teams tend to be small, one person's move can make a big difference.

Imitation
If a fund does spectacularly well, other managers will copy it and perhaps outdo it.

Fund size
If a fund manager is particularly successful, investors tend to move their money into that fund. Suddenly the fund is bigger and may be harder to administer. Suddenly, too, the manager can't grow all the current share holdings equally, because some shares just aren't available in large numbers. This is a particular problem in New Zealand.

For these reasons, failure may follow success. Experts advise staying away from any fund that is growing fast. On the other hand, it also pays to be wary of small funds, which are fairly common in New Zealand. They can become too small to be economically viable, and are more likely to be taken over by larger funds, leaving investors under entirely new management.

If only we knew

A while back, the Economist magazine reviewed asset performance over the 20th century.

The researchers looked at a hypothetical investor who was perfect at forecasting. He invested $1 at the beginning of 1900 into what was to be the best-performing asset type in that year.

Then, at the start of every year, he moved his money into what was to be that year's winner. By 2000 he would have had $1300 and 12 more zeros!

It just goes to show the power of compounding high returns over very long periods.

But if instead he had put his money into the previous year's best performer - something many people are tempted to do - his $1 would have grown to a mere $290 - with no zeros.

* During September, we are running excerpts from Mary Holm's book 'Get Rich Slow'.