Let's start by considering a naïve forecaster - one who knows past returns but nothing about the future. He would think as follows.
The long-term average real return on equities has been 5.6 per cent per year since 1900. Because I know nothing about what'll happen in 2015, my least bad guess must be that it will be an average year, so I should expect a 5.6 per cent real return. With economists expecting 1.9 per cent inflation this year, this implies a total return of 7.5 per cent. For the FTSE 100, 3.5 percentage points of this will come from the dividend yield. This implies a 4 per cent price rise, which would leave the index at 6874 by the end of 2015.
But of course, there's uncertainty around this. The standard deviation of real returns since 1900 has been 21.1 percentage points. If we assume returns are normally distributed, this implies that there's a two-thirds chance of the index ending the year somewhere between 5423 and 8324, with a one-in-six chance of it being below that range and a one-in-six chance of it being above it.
Now, the assumption that returns are normally distributed isn't exactly right. But that doesn't overturn the main point - which is that there is, from a naïve forecaster's point of view, enormous uncertainty about returns.
The question, therefore, is: can we reduce this uncertainty? We can certainly do so by self-deception - by persuading ourselves that some ju-ju man has privileged insight into the future. A more rational way to do so, though, would be to look for indicators that have some ability to predict returns.
One good candidate here is the dividend yield. Between 1996 and 2014, it explained 27 per cent of the subsequent variation in annual changes in the All-Share index – although this leaves 73 per cent unexplained.
Right now, the dividend yield predicts that the All-share index will rise 7.3 per cent this year. If the FTSE 100 rises as much, this would take the index to 7092. However, there's still massive uncertainty around this. 1996-2014 relationships imply that there's around a two-thirds chance of the FTSE 100 ending the year between 6134 and 8049. This doesn't seem very helpful.
However, there's one indicator that's done even better at predicting returns: foreign buying of US equities. The idea here is that such buying is a sign of global investors' confidence. High buying is a sign of high confidence. But this often means investors are over-confident and hence that global share prices are too high and that they will fall. Foreign buying peaked in 2000 - just before the tech bubble burst - and again in 2007, just before the financial crisis. Foreign buying of US equities alone (measured over a 12-month period) explained over four-fifths of the subsequent variation in annual changes in the All-Share index between 1996 and 2014.
Better still, combining foreign buying and the dividend yield gives us an indicator that has explained almost half the variation in annual returns since 1996.
Because foreigners were net sellers of US equities in much of 2014, this indicator now points to big rises in share prices. If post-1996 relationships continue to hold, UK equities will rise 28.6 per cent this year, implying that there's around a two-thirds chance of the FTSE 100 ending the year between 7531 and 9470.
This sounds fantastic. But there is a catch. Equity returns last year were well below what foreign buying of US shares predicted. This might be a sign that the once-stable relationship between such buying and subsequent returns has broken down. This would have happened if foreigners sold US shares not because they were pessimistic about global equities but rather because they simply rebalanced their portfolios away from US stocks and towards other risky assets. And it might be that the most recent selling of US shares - foreigners dumped $27bn of them in October - tells us not that confidence is unreasonably low but rather that oil exporters have suffered falling revenues and so have had less cash to invest.
This takes us back to where we started. Maybe the best prediction we can make is for equities to see single-digit positive returns, with a high chance of something much worse or much better.
If you think this isn't very helpful, you'd be right. But the problem here is not a failure of the economics profession. Instead, it's that returns are inherently unpredictable. Imagine there were some indicator which did predict big returns. Investors would have discovered this, and so bought equities. In doing so, they would have raised prices to a level from which subsequent returns were less spectacular. This might have been the fate of our foreign buying indicator. This tells us that returns can only be predictable if the source of that predictability remains unknown to other investors. To say the least, this is unlikely.
None of this means that economists and financial advisers are useless. Instead, their function is to show us how to manage risk, and to warn us against the many errors of judgment that can cost us money. Insight into the future might – just might – be possible, but it is not to be relied upon.
