Some myths die hard. One of these is the idea that older people should own fewer equities than younger ones. In many cases, this is not true.
The reason for believing it is that older people have shorter time horizons than younger ones and therefore less time to recoup losses. Both the premise and inference here are dubious.
The premise is doubtful because many older folk do have long time horizons. Someone in their 60s can reasonably expect to live another 25 years. And if you’re planning to leave a bequest to your children or grandchildren your time horizon exceeds your life expectancy.
The inference is also mistaken in a more interesting way. It seems to have a grain of truth. If you toss a coin once and call heads you have a 50 per cent chance of losing. But if you toss it enough times you’ll be right around half the time. Over time, luck cancels out. The same can be true of stock markets.
Let’s put some numbers on this. We have data on share prices since 1700. During this time, they have risen by an average of 1.1 per cent per year in real terms: if this seems low, remember that most returns come from dividends rather than price appreciation. The standard deviation around this has been 15.3 percentage points. If returns in one year are independent of those in the next, then this standard deviation rises with the square root of time. This implies you have a 47 per cent chance of losing money in one year, but only a 36 per cent chance of losing over 20 years. Your chance of a loss is thus smaller in the long run.
This, however, is no justification for regarding shares as safer in the long run. Although you have a smaller chance of losing in the long run, you have a bigger chance of really horrible losses. In one year, the financial crisis of 2008 will happen only once. But over 20 years it could happen several times. Yes, you’ve less chance of losing money in the long run, but you’ve more chance of a nastier loss. For many people, these two risks cancel out, making shares as attractive in the short run as in the long.
You might object here that I’m making an unjustified assumption – that returns are independent from one year to the next. If, however, returns mean-revert – with losses more likely to be followed by gains – then the market is indeed safer in the long run and so the young should own more shares than the old.
My chart tests this. It shows the actual volatility of returns over different horizons compared with the volatility we’d expect if returns were serially independent. If returns really were independent, the two lines would be the same.
Source: Bank of England
But they are not. Returns over six or seven years have been slightly more stable than serial independence implies, but returns over 15 years or more have actually seen more volatility. In this sense, equities are riskier for long-term investors than for short-term investors. (My data is drawn from returns since 1700, but the story is very similar if we look only at those since 1900 instead.)
There are more historic examples of this than you might think. In real terms, shares did not return to their 1720 level (the time of the South Sea bubble) until 1850. They were lower in real terms in the 1980s than they were in 1929. And of course they are lower now than they were in 2000. If you spend your dividends, you can lose a lot over the long term.
In fact, this might actually understate the long-term risks we face now. I’ve recovered from every illness I ever had or feared I had. But this doesn’t mean I’m immortal. Similarly, the fact that shares have recovered from so many past mishaps and fears (recessions, financial crises, inflation, fear of communism and so on) does not mean they’ll continue to do so. Philippe Jorion and William Goetzmann have estimated that a third of the stock markets that existed in the 1920s subsequently suffered a long-term closure that wiped investors out. For them, equities were a better short-term investment than long-term one. Put it this way. We can be pretty confident that most equities will still exist next year. Can we be so sure they will in 30 years’ time?
If shares are riskier and more uncertain in the long term, then the conventional advice that people with shorter horizons should hold fewer of them is plain wrong. Older people should not, therefore, reduce their equity exposure.
Or should they? There is in fact one good reason why some of them should do so. It was pointed out back in 1996 by Ravi Jagannathan and Naryana Kocherlakota. When you are working you can offset stock market losses by saving out of your salary. When you retire you lose this diversification and so equities become riskier, and so you should hold fewer of them.
This, however, doesn’t apply to everyone. If you retire on a final-salary pension you’re swapping one diversifier for another: you pension is equivalent to a big holding of bonds. That means you can stay invested in equities, except to the extent your pension is lower than your salary. And if your salary is sensitive to macroeconomic conditions – if you’re a banker or architect, say, rather than a doctor – then it is a poor way to diversify equity risk. For you, shares might actually become safer when you retire, so you should hold more.
You should therefore forget the old advice that you should hold fewer shares as you get older. It’s an old wives’ tale. Instead, you should ask the same questions at any age: do expected returns on equities compensate for their risk? How well am I able to take equity risk? For many of us investors, the cliché is true: age is just a number.
