The Jonathan Davis column
It is a brave man who is prepared to stick his head above the parapet and try to predict the future course of market returns. For anyone trying to do some serious long-term investment planning, however, the exercise must be undertaken. The most important criterion to adopt is to make sure that the assumptions you make are: (i) consistent with each other; and (ii) based on realism.

I make no excuse, therefore, for returning to the Barclays Capital gilt- equity study mentioned last week. As with equities generally, which reward those who take the risk of owning them with handsome returns over time, the author of the Barclays study, its markets and strategy expert Michael Hughes, is well paid for sticking his neck out and predicting how the next 10 years might turn out.

So I make no apology for putting Michael's neck on the block again in public. A year ago he correctly made the point that while the stock market looked overvalued on all conventional valuation models, that in itself was no reason to believe that it would not continue to do well at least for a time.

At the time, he was criticised in some quarters for his bullish view, partly on the grounds that as he worked for a broking firm (BZW), he would be bullish about equities, wouldn't he? Fair enough, I suppose, but as we now know, it turned out to be an excellent call, with 1997 another starring year for both shares and gilts. This year, with Barclays having sold BZW, at least the accusation that Michael might be talking the book of his equity salesmen is one that can no longer be reasonably held against him!

Yet, as I indicated last week, he is still in fundamentally optimistic vein. He believes, as do I, that gilts will continue to become relatively more attractive as an investment class: their rerating vis-a-vis equities will continue. He sees yields on gilts reverting to something closer to their level before the great inflation horrors of the 1970s (5-6 per cent), and the yield ratio (the yield on gilts as a multiple of the yield on shares) falling from its current level (around 2.0) to 1.5 or less.

He projects gilts producing a total return of 8.5 per cent a year over the next 10 years - less in nominal terms than the last two decades, but a still handsome 6.0 per cent a year in real terms. This is a powerful argument for including a higher proportion of gilts in any investor's portfolio today than would have been sensible in the past 20 years. As the table shows, it is the assumptions about equity returns which look more suspect. The dividend yield on the market at the start of this year (3.3 per cent) is lower than at the start of all the previous 10-year periods covered by the table.

History suggests that a dividend yield in this range is likely to be followed by a period of negative real returns over five years and only a marginally positive real return over the next 10 years. Yet the Barclays Capital study suggests that equities are capable of continuing to appreciate in value by 7.0 per cent a year and generate a total return (including dividends) of 10.5 per cent a year.

That certainly looks ambitious and requires special circumstances to justify it. One obvious one is the secular decline in inflation that we have seen since the mid 1980s. In real terms, if you project inflation of 2.5 per cent a year, as Barclays does, then 10.5 per cent a year equity returns are not out of line with past experience. It equates to real capital appreciation of 4.5 per cent per annum (7.0 per cent minus 2.5 per cent) - which is actually fractionally lower than the post-war average of 4.6 per cent per annum. Another underlying assumption you need to arrive at such an outcome is that companies will continue to be able to generate and retain the much higher proportion of GDP that they have achieved so far this decade.

Michael Hughes admits that this may be optimistic, but argues that it is not impossible in the current environment of low inflation and increasing globalisation. He also has some interesting statistics about the age profile of the country. As we all know from the "pensions timebomb", the proportion of the population aged over 55 is set to reach a record level in the next 15 years. If past precedent is any guide, this should result in a fundamental shift in the growth of savings, which in turn could provide a fundamental shift in the valuation of both gilts and equities.

Put all these forces together, and what you have, conceivably, is a set of circumstances which could justify a continued period of good performance from shares, alongside a relative return to favour by gilts. Of course, there are a lots of ifs and buts - war, inflation, deflation are all threats. So too is the possibility that there will come a political backlash against large company profitability from tax-starved governments. Michael Hughes' point is that within the constraints of the long run historical averages, valuation parameters can and do change for quite long periods of time.

We need to be alive, he argues, to the possibility that the era we are facing is a genuinely unprecedented one, in which the savings rate soars, corporate profits remain strong and inflation is restrained for another decade. Too good to be true? One of the defining characteristics of the top of a bull market, goes an old saying, is its ability to "draw in higher intellects". My view remains that it is more prudent to assume a less rosy outcome and hope to be pleasantly surprised rather than aim too high and be brought crashing down. The message is: stay invested in equities, but don't go overboard.