Equities, we are told, should be a long-term investment. Recent evidence, however, suggests that they are not.
My chart shows this. Each point on the line shows the annualised return you would have made on the All-Share index relative to gilts (including dividends), had you bought on the date on the horizontal axis. So, for example, the one percentage point for December 2004 means you would have made one percentage point a year more on equities than gilts if you had bought in that month.
This reveals two things.
One is that even over long periods equities have underperformed gilts. If you’d bought in January 1997 you would have made 5.8 per cent a year from equities, but 6.3 per cent from gilts. That’s 24 years of equity underperformance – more than half of a working lifetime.
Secondly, timing matters. If you’d bought equities in the late 1990s or 2006-07, you would have underperformed for many years. But if you’d bought in 2003 or 2009, you’d have made a decent equity premium. You cannot ignore market timing in the hope that equities will come good in the long run.
We can think of this another way. Since January 2010 UK equities have doubled your money if you had reinvested dividends. That’s respectable enough. All of this return, however, has come in only the best 10 months since then. Fewer than eight per cent of months account for all of the market’s returns in the past 11 years.
The All-Share index made us more money last week than in the entire previous three years. Which reminds us that returns are skewed.
What’s true of the aggregate market is even more true of some particular strategies. My portfolios of negative momentum and high beta stocks have both made more than 40 per cent since October. But both have still lost money since 2011. Buying dog stocks or high beta ones can therefore be a good short-term play but a lousy long-term one.
So, should we forget the conventional advice and think of shares as only a short-term investment?
No. The reason to think of equities as a long-term investment is to be found not in recent data but in economic theory or common sense. Sometimes, the past is a poor guide to the future.
Quite simply, equities should outperform gilts or cash over time because they are risky and so should deliver a risk premium on average over the long run. That this hasn’t happened since the 1990s is because gilts have done extraordinarily and unexpectedly well as yields have trended down. But we’ve no compelling reason to expect this trend to continue. And if yields stay around current levels, we should see equities outperform gilts on average over the long run.
Conventional theory, however, predicts only a small equity premium. Which means that even over longish periods there’s a non-negligible chance of equities under-performing safe assets. A two percentage point annualised premium implies that there’s around a one-in-four chance of equities under-performing even over 20 years simply because the market is volatile and bad luck might not even out over such a period.
There is, though, another reason why equities should do well over the long-run. It’s that regular dividend income compounds nicely if you reinvest it. Put it this way. Adjusted for inflation, the All-Share index is lower than it was in 1996. That’s a quarter of a century of real zero capital gains. If we add in reinvested dividends, however, the market has delivered a real return of over 3 per cent a year. That’s only marginally better than gilts, but it’s reasonable.
Again, though, we must be careful here. It’s not just dividend income that compounds over time. So do fund managers’ fees. An extra half percentage point of charges levied on a 3.5 per cent annual return will cost you almost £2,000 for each £10,000 you invest. There’s a reason why fund managers think you should be in equities for the long term.
Even so, they might be right. There’s another reason why. Yes, shares’ long-term returns tend to be bunched into just a few months. But we cannot predict what months these will be. So we should hold them for the long term to ensure that we get such months – in the same way we hold premium bonds for years in the hope of getting a big prize some time.
Although we cannot predict short-term returns, history suggests we can predict longer-term ones simply by looking at the dividend yield: a high yield predicts high returns and a low yield low returns. Right now, the yield is around its average and so is pointing to around-average returns. Average might not be very good. But it’s good enough to regard shares as a long-term investment.
