History repeats, mistakes and all

Sharemarket crashes are a fact of life, but it's trying to beat them that causes the real damage, writes PAUL DYER.*

With share investments, volatility is a fact of life.

Even the severe fall in global sharemarkets over the past year is similar in size and duration to those that have typically occurred every five to 10 years.

This time the fall was unusual only in that it was a long time since the last one in 1987.

Similar falls have happened often in the past and will do so in future.

This does not mean investors should abandon sharemarkets.

Why not? Because history teaches us that entering markets after they have risen, or getting out after they have fallen, is a route to certain failure.

In part this is because the strongest sharemarket returns have often followed large declines.

The evidence is that most investors are poor at timing markets, and conceivably reduce their long-term returns by half or more through reacting to recent moves.

This experience may be repeating in 2001, with the flow of money into share investments drying up just as major sharemarkets look their most attractive for several years.

In contrast, there are now strong inflows to bond markets whose valuation look very stretched.

By far the best advice to investors remains to choose an appropriate well-diversified investment mix and to stick to it.

While sharemarket volatility can't be avoided, unfortunately the consequences are often magnified by investors' tendency to buy into markets that have already gone up, and to get out of those that have already fallen.

The clearest evidence of this can be seen in the United States, where we have good quality monthly data about investment flows back to 1984. The evidence from New Zealand, the UK, Europe and Australia all follows the same pattern.

Studying the amounts flowing into US sharemarket mutual funds (the equivalent of unit trusts) reveals that there were strong inflows before the crash of 1987 and throughout the bull run of 1995-2000.

By the first quarter of last year, inflows were at an all-time high.

By contrast, inflows slumped following periods of market weakness such as after the 1987 crash and during the setbacks associated with the 1990-91 Gulf crisis and the 1998 Russian crisis.

This year has seen the first consecutive monthly outflows in over a decade.

The problem is that these flows tend to happen at precisely the wrong times. Not only do investors who buy late miss the gains from rising markets, but they tend to invest at peaks and sell at troughs.

There is in fact a negative relationship between the amount of money flowing into sharemarket funds and market returns over the following 12 months.

How much does this tendency cost investors? The answer is a lot. Two examples show that investors conceivably halve their returns because of this.

1: First, consider a hypothetical investor who chooses every year whether to hold shares or keep their money in the bank, depending on which approach had produced the higher return over the previous year.

Again using US data, this investor would have held cash rather than shares in 17 out of the last 55 years.

The real (after inflation) return received over the entire period would have been 5.1 per cent a year.

But by keeping their money continuously in the sharemarket, the investor would have earned 10 per cent a year.

More importantly, for those 17 years when the investor was out of the market, shares returned an average 17.4 per cent a year, significantly more than the long term average level.

2: But is this what investors actually do? Using data for the inflows into US mutual funds, we calculated the cumulative returns to an investor who each month put money into the US sharemarket in proportion to actual mutual fund inflows - more following periods of strong sharemarket returns, less following periods of weak growth, or falls in the market.

We looked at the returns generated by this approach compared with the alternative strategy of contributing the same total amount by buying shares regularly each month.

Investing the same amount each month produced returns that were almost 50 per cent higher.

On average, US investors have received only about half the returns they could have if they had taken a more disciplined approach. The same will hold true for all other sharemarkets.

Hindsight investing is costly because of the tendency for high sharemarket returns to follow periods of low ones (and vice versa).

For example, it may come as a surprise that the strongest five-year period of returns for the US sharemarket over the past 200 years was during the great depression of the 1930s.

Between May 1932 and May 1937, US shares returned 367 per cent. This was a partial rebound from the decimation of share prices between 1929 and 1932.

The point is that investors who sold out near the bottom had double reason for regret.

In fact, for the decade to the start of 1937, US share prices more than doubled, despite falling over 80 per cent between 1929 and 1932. As always it was the johnny-come-latelies, those who entered and exited the market at the wrong times, who suffered.

It gets worse. Bad as these examples are, actual investor behaviour reduces average returns still further.

We treated "the sharemarket" as a single entity, which clearly it is not. In real life, incoming flows tend to be directed towards those sectors which have recorded the highest recent returns.

A recent example is the growth vs value debacle.

"Growth" stocks are those companies expected to have higher than average growth, which therefore sell at higher prices, relative to their earnings. "Value" stocks are the remainder of the market, companies which sell at lower prices because of lower growth prospects.

From 1996 to 1999 growth stocks, which include almost all of the technology sector, strongly outperformed value stocks.

Not only did total inflows into the market balloon, but new funds were increasingly diverted into growth sectors. In 1999 growth fund inflows were more than 100 per cent of total inflows. In other words, people were withdrawing money from value-based funds.

The number of growth-oriented funds mushroomed. By contrast, many value funds wound up and some value managers quit the industry.

Since March 2000 growth stocks have underperformed value stocks by over 60 per cent. In other words, investor flows though 1999 and early 2000 were completely misdirected.

Our calculation of the costs of hindsight investing would be much greater if these aspects could be included.

We have also looked at inflows to US bond market funds, and find the same pattern. Inflows strongly follow past returns, and are negatively correlated with future returns.

In early 2000, when - in hindsight - fixed interest offered excellent value, there were strong outflows. This year there are strong inflows, even though average Government yields globally are now less than 4 per cent.

The lesson: hindsight investing simply doesn't work.

There should be a strong presumption against reducing the proportion of your investments in shares simply because markets have fallen. The same applies to buying into rising markets.

The risk is that this experience may be repeating in 2001.

Sharemarket inflows have been weak everywhere this year in response to falling markets. Inflows to fixed-interest investments have been strong, driving yields down everywhere.

Very likely investors are making their usual mistake of getting out of markets just as prospects are brightening.

* Paul Dyer is head of investment strategy, AMP Henderson Global Investors (NZ). This is an edited version of a longer article on the lessons investors can learn from the recent fall on sharemarkets.