Hamish McRae: Equities have recovered but don't expect double-digit returns as well

When looking forward it is best to start by looking back. In the past few days two of the three annual surveys of long-term investment performance have just been published, from Credit Suisse First Boston and ABN-AMRO in conjunction with the London Business School. The third, from Barclays Capital, comes out later this month.

Most people will be aware that the rise in shares last year on all the main markets ended three years of decline, itself something that had not happened in Britain for more than 50 years. The scale of the decline in the UK and the United States was the greatest since the early 1970s crash.

If you look at shares world-wide, according to the LBS team, the three-year fall was the greatest since the 1930s. So we have come through a very unusual experience and one that ought, on any historical view, to be well behind us. That is not to say that shares go up from here. But anyone considering pension fund performance, or the case for holding equities rather than bank deposits, should be clear that we have just come through a once-in-a-generation equity hurricane. On past form the next such hurricane will come some time between 2030 and 2060 and I think we can let someone else worry about that. What might happen to fixed interest securities is another matter and we'll come to that later.

The first point that leaps out, looking at the data for last year, is the way all the major share markets moved pretty much together. Take the five biggest markets, the US, UK, Japanese, French and German ones. The bottom was in March, at the time of the Iraq invasion, and the top for all but Japan was around year-end. Japan peaked in October. The scale of the rise differed, with the total return (i.e. share price movement plus dividends) from Germany climbing by 38 per cent and the UK only 21 per cent. But the fundamental shape of the movements are much the same.

The best explanation I can see for that is the dominance of the US market. With the partial exception of Japan, we take our lead from America. But then look at the scale of US market capitalisation vis-à-vis the rest of the world. As the pie chart demonstrates, at the end of last year the US accounted for 53 per cent of world equity markets. The UK was a poor second with 10 per cent, just ahead of Japan, and well ahead of France and Germany with 4 per cent and 3 per cent. No other country accounted from more than 2 per cent of the world market. This is an American game.

It is also a developed world game in the sense that developing countries account for less than 6 per cent of total market capitalisation. Countries such as China and India may be becoming economic giants but in investment terms they remain tiny. Incidentally, there also does not seem to be much connection between economic progress and financial performance: the Shanghai market is depressed, while the Bombay one is booming.

The second graph shows UK real equity returns going back to 1869, a useful point because it predates the 1870 start of what came to be known as "the great depression". That was depression of agricultural land prices and prices in general, rather than a depression of economic activity. Over this long period the real return from shares has been 5.7 per cent. This is pretty good, except that it does not count any effect from taxation or trading costs. Accordingly an ordinary investor would be pushed to achieve this.

On that very long view, share prices are now well within the "normal" limits of one standard deviation either side of the mean. But they are not screamingly cheap. The overvaluation three years ago was the greatest it had ever been in the previous 130 years. Even last March, at the bottom of this market cycle, they were a touch above their long-term trend. On the other hand, if you believe that we are in for a long period of deflation akin to 1870-1914, there could still be a decent run for a few more years. Look at share performance in the 1880s and 1890s.

But there must be concerns. For myself, the principal one is the extent to which past performance has relied of dividends rather than capital growth. We tax dividends very highly now, even if they are held in pension funds. So it is hard to see how they can do much to support the present level of prices, particularly since the price-earnings ratio of London, at about 18, is still above its long-run average for the last 75 years of about 15.

Taxing dividends is a bit of an anomaly, particularly since interest on bonds counts as an expense that companies can set against tax. So there is a huge incentive for companies to raise capital by issuing more debt rather than by issuing more shares. The US administration does propose to cut tax on dividends and I expect this will eventually become the new global standard. But it is not going to happen here for a while.

The trend in dividends over the past century is shown in the bottom graph. Here there is a distinction between the US and the UK. In the States there has been a long-term downward trend, whereas here there has been no real trend at all. Clearly the collapse of the US dividend yield during the late 1990s to just 1 per cent was unprecedented. But note that it has not recovered much. In the UK we are pretty much back to the long-term band of 3.5-5.5 per cent, so dividends so give much more support to UK shares than US ones. If there is a bad patch in the next couple of year, though, do not expect the UK tail to wag the US dog.

So should we rebalance portfolios towards fixed-interest securities? One of the aspects of the current regime of investment protection that I find most absurd is the compulsory ranking of portfolios that are mostly in equities as "high risk" while those skewed to gilts are seen as a lower risk. At present levels, any fixed-interest portfolio is surely a higher risk than an equity one for the simple reason that fixed-interest securities are even more over-valued than shares.

If shares were overvalued by about 80 per cent at the end of 1999 and are about 20 per cent overvalued now, gilts are currently 60 per cent above their long-term trend. CSFB notes these points and concludes: "Perhaps gilts are more expensive than they might seem at first glance." It also notes that the institutional investors, having unloaded much of their equity portfolio and switched to fixed-interest, are now moving back.

Any rise in inflation would be disastrous for bonds. My own observation here is that so much money has been pumped into the world economy by the central banks, particularly the Federal Reserve, that even those of us who think long-term trend is for deflation, a modest kick up in global inflation is likely this autumn or next year. If that is right, bonds are dangerous territory indeed.

Perhaps the most useful facet of this whole long-term perspective is the way it makes us realise that the double-digit returns of the 1980s and 1990s were utterly abnormal. The surge in inflation of the 1970s was abnormal and as that was painfully corrected, shares were driven from being very undervalued to being very overvalued. In 1974 they were as undervalued as at any stage since the First World War (middle graph again) and at the end of 1999 they were more overvalued than at any stage since 1869. Ultimately share values are likely to grow at much the same rate as GDP - maybe a little faster in the UK because our companies have large international activities and the world economy will grow a touch faster than our own one. That suggests that the long-term rate of growth in returns of around 6 per cent is a good working assumption - not the 12 per cent or more that most major markets experienced in the 1990s.