Secrets Of Success: Statistics prove history is not bunk

Here are some more interesting findings from the just-published "Global Investment Returns Yearbook" from the London Business School and ABN-Amro, with comments from me. As Mark Twain used to say, history never repeats itself, but it does rhyme. There is plenty to ponder in the long-run figures of different investment asset classes.

First, equities provide the best long-run returns. This is generally true - and as it should be, as they are riskier than other asset classes. Nevertheless, the argument for the superiority of equity returns is heavily dependent on US (and to a lesser extent UK) experience. Many markets have had different experiences at different times.

For example, in the century and a bit since 1900, the longest period of time that a United States investor has had to wait for the stock market to produce a positive real return (that is, after allowing for inflation) has been 16 years. In the UK, the comparable figure is 22 years. Yet in France, Germany and Japan, the figure is more than 50 years.

It is true that these long periods of long-run disappointment all took place in the first half of the last century, a time of two world wars, hyperinflation in some countries and (so many historians would argue) a period when central bankers were not as sophisticated as they are today in applying monetary policy.

Nevertheless, history tells us that bad things do happen from time to time, and stock market investors should not blithely assume that long-run superior returns from equities are somehow guaranteed.

Second, a pattern of boom and bust in equity markets has been the norm. The post-2000 bear market was actually one of the worst in magnitude in the record. According to the LBS/ABN-Amro database, the world stock market fell by 44 per cent in real terms from 2000 to the end of 2002.

This was similar to what happened in the 1973-74 oil shock, and only slightly less dramatic than the 1929-31 Wall Street Crash, when stock markets declined 54 per cent in real terms. (Note that these figures are based on calendar years: the actual peak-to-trough decline in both cases was somewhat larger, so felt even scarier at the time.)

All three bear market episodes followed decades in which the stock market had been exceptionally strong in historical terms, at least in the US and UK. World stock markets rose by 200 per cent in real terms in the 1920s, for example, by 500 per cent in the 1950s and by nearly 400 per cent in the 1980s and 1990s combined.

In leading economies, the most dramatic gains of all time came from the 1950s rebuilding of Germany, when the German market delivered a 4,000 per cent real gain. Although Harold Macmillan said Britons had never had it so good in the 1950s, the UK market in fact lagged way behind the world as a whole.

One of the things that marks out the 2000-03 bear market is that it is virtually the only example of a market crash that was almost entirely driven by overvaluation rather than external events. The 1920s, with similar speculative fervour in the years up to the crash, is the closest example.

The fact that the world has suffered relatively minor economic consequences this time round, compared to the horrors of the 1930s slump, is in part a tribute to the skills of Alan Greenspan, the Federal Reserve chairman (though we don't yet know whether his cheap-money policies will have further adverse consequences down the line).

The third observation is that diversification pays. It does. While several countries have delivered higher returns over the past 105 years than an aggregated world market index, the volatility of the world market is lower than any of the 17 leading markets whose records go back to 1900. The worst single-year return from the world market (minus 34 per cent in 1931) is also lower than that of any individual country.

Which country has the best risk-reward ratio on this measure? Australia, with an annual real return of 7.7 per cent with volatility of 17.6 per cent, against the world market's 5.7 per cent return and 17.2 per cent volatility. The worst markets have been Germany and Italy (though in the former case the figures are dragged down by the hyperinflation of the 1920s).

Fourth, bond markets are notoriously vulnerable to inflation. They are. The long-run compound real rate of return from government bonds has been 1.6 per cent, against 2.5 per cent for Swiss bonds, 1.9 per cent for US bonds and 1.4 per cent for UK bonds, but the experience has been very different in times of low and high inflation. Several other countries, including Germany and Japan, have negative long-run rates of return on bonds.

One consequence of these figures is that investors who are used to demanding a 3 per cent premium over inflation to buy gilts may be being unduly cautious. If (and it is a big if) central bankers have indeed learnt how to control inflation effectively, we should all be getting comfortable with buying bonds on lower premiums. This has in fact been happening in recent years.

Finally, value and small-company shares have produced the best long-run returns historically. The strong performance of both in the last five years has only reinforced this long-run trend. LBS/ABN-Amro says the comparable annual rates of return for the UK market over the past 50 years are: equity market in general 13.6 per cent a year; small-cap index 16.4 per cent a year; and micro-caps 20.1 per cent a year.

The comparable figures for high- and low-yielding stocks (which the professors take as a proxy for value against growth) are: low-yielding stocks 8.0 per cent a year; equity market in general 9.7 per cent a year; and high-yielding stocks 11.3 per cent a year.

History on this occasion does not, unfortunately, tell us much of use about how to rotate between these styles, except that in the long run they are the way to bet. In the case of small caps, which are riskier, it is important to take into account the impact of transaction costs, which are invariably higher the smaller the company.

Large caps did exceptionally well in the 1990s, and growth was strongly in favour at that time, which is why Vodafone came to account for 14 per cent of the UK index at one point.

The past five years have redressed the balance to some extent, but for how long? Only Rip Van Winkle investors - those who buy a portfolio and leave it to mature for a decade, like claret - are spared the need to confront these style issues from time to time. And we are approaching one such potential turning point now.

jd@intelligent-investor.co.uk